oversight

Tennessee Valley Authority: Future Study of Lake Levels Should Involve Public and Consider Costs and Benefits

Published by the Government Accountability Office on 1999-05-17.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                  United States General Accounting Office

GAO               Report to the Honorable
                  Van Hilleary, House of Representatives



May 1999
                  TENNESSEE VALLEY
                  AUTHORITY
                  Future Study of Lake
                  Levels Should Involve
                  Public and Consider
                  Costs and Benefits




GAO/RCED-99-154
      United States
GAO   General Accounting Office
      Washington, D.C. 20548

      Resources, Community, and
      Economic Development Division

      B-282488

      May 17, 1999

      The Honorable Van Hilleary
      House of Representatives

      Dear Mr. Hilleary:

      As requested, this report addresses how the Tennessee Valley Authority (TVA) manages and
      operates its multipurpose tributary projects—consisting of dams and lakes on the tributaries of
      the Tennessee River—for various purposes, such as flood control, navigation, hydroelectric
      power production, and recreation. Our report provides information and analyses on the
      purposes served by these projects, operational changes made to these projects in 1990, actions
      taken by TVA since 1990 to address requests for changes in project operations, and TVA’s plans
      for any future changes in project operations. This report also provides a recommendation to the
      Chairman of TVA’s Board of Directors.

      As arranged with your office, unless you publicly announce its contents earlier, we plan no
      further distribution of this report until 30 days after the date of this letter. At that time, we will
      provide copies of the report to appropriate House and Senate Committees; interested members
      of Congress; Craven Crowell, Chairman, TVA’s Board of Directors; The Honorable Jacob Lew,
      Director, Office of Management and Budget; and other interested parties. We will also make
      copies available to others on request.

      Please call me at (202) 512-3841 if you or your staff have any questions. Major contributors to
      this report are listed in appendix VI.

      Sincerely yours,




      Susan Kladiva
      Associate Director, Energy,
        Resources, and Science Issues
Executive Summary


             The Tennessee Valley Authority (TVA), a wholly owned government
Purpose      corporation, is, by some measures, the nation’s largest electric power
             producer. TVA is also responsible for the use, conservation, and
             development of the natural resources related to the Tennessee River. In
             carrying out these responsibilities, TVA relies on an integrated system of 54
             dams and associated components—referred to as projects—to harness the
             Tennessee River and its tributaries to serve various purposes, such as
             navigation and flood control, for the public’s benefit. Thirteen of these
             projects, referred to as multipurpose tributary projects, consist of dams
             and lakes on the tributaries to the Tennessee River, such as the Douglas
             and Cherokee projects, that provide multiple public benefits. Concerns
             have been raised about TVA’s operation of these multipurpose tributary
             projects to satisfy the multiple and often competing public benefits.

             Representative Van Hilleary asked GAO to provide information on (1) the
             purposes served by TVA’s multipurpose tributary projects and how TVA
             operates these projects within its integrated system, (2) the operational
             changes TVA made to these projects as a result of its December 1990 review
             of its project operations and the major factors influencing these changes,
             (3) the actions TVA has taken since the 1990 review to address requests for
             changes in the way it operates these projects, and (4) TVA’s plans for any
             future changes in the way it operates these projects. GAO is also providing
             information on a selected update by TVA of its analysis performed in the
             1990 review.


             TVA was established by the Tennessee Valley Authority Act of 1933 to,
Background   among other things, improve the navigability and control flooding of the
             Tennessee River and provide for the Valley’s agricultural and industrial
             development. The generation and transmission of hydroelectric power was
             established by law as an additional benefit secondary only to navigation
             and flood control. To carry out its river management responsibilities, TVA
             constructed or acquired its 54 projects to serve many purposes. These
             projects have a dam and can include several components, such as a lake
             behind the dam and hydroelectric power facilities at the dam. Some of
             these projects were constructed for specific purposes, such as producing
             hydroelectric power. Other projects, such as the 13 multipurpose tributary
             projects, were constructed for multiple purposes, such as producing
             hydroelectric power and helping control flooding.

             TVA has guidelines, which it refers to as lake level policies, that prescribe
             the levels that must be maintained at the lakes at various times during the



             Page 2                       GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                   Executive Summary




                   year. Under these guidelines, most of the multipurpose tributary projects
                   are subject to significant lake level changes during the year as a result of
                   TVA’s flood control and hydroelectric power production efforts. Annually,
                   TVA lowers the lake levels at these projects during the late summer and fall
                   in order to provide space for the rainfall and runoff that occurs in the
                   winter and early spring. According to TVA, while large storms can occur
                   throughout the year, the major regional floods on the Tennessee River
                   normally occur between December and April. As TVA lowers the lake
                   levels, a process referred to as the “drawdown” of the lakes, TVA
                   generates hydroelectric power with the water released from the dams. As
                   the flood risk diminishes in the spring, TVA allows the lake levels to rise
                   from rainfall and runoff flowing into the lake in order to reach desirable
                   summer levels for recreational purposes.

                   TVA has performed studies and reviews in the past examining whether
                   changes in lake levels could be made to improve recreational uses of TVA
                   lakes. The most recent review was published in December 1990, which
                   resulted in TVA delaying the annual lake-level drawdown at multipurpose
                   tributary projects from Memorial Day to August 1. This delayed drawdown
                   allowed lake levels at these projects to remain higher during the summer
                   recreation season.


                   The operation of the multipurpose tributary projects1 serves several
Results in Brief   purposes—primarily navigation and flood control, and to the extent
                   consistent with these purposes, hydroelectric power production. These
                   three operating priorities are contained in the TVA Act. TVA can permit
                   operation of the projects for other purposes, such as recreation and water
                   quality, subject to the three statutory purposes. In operating its integrated
                   system, TVA often finds that the multiple purposes served by the projects
                   can conflict and/or compete with each other. As a result, TVA attempts to
                   balance the various purposes served to provide the greatest public
                   benefits from the waters of the Tennessee River and its tributaries while
                   adhering to the operating priorities.

                   A key change resulting from its December 1990 review of project
                   operations was TVA’s delaying the annual lake drawdown at the
                   multipurpose tributary projects from Memorial Day to August 1. The major
                   factor influencing this change was that, of the seven alternatives examined

                   1
                    For this report, GAO defines multipurpose tributary projects as including 14 projects—the 13 projects
                   mentioned (Boone, Chatuge, Cherokee, Douglas, Fontana, Hiwassee, Melton Hill, Norris, Nottely,
                   South Holston, Tellico, Tims Ford, and Watauga)—and another project (Blue Ridge) because this
                   project has an annual drawdown cycle similar to most of the other 13 projects.



                   Page 3                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                           Executive Summary




                           in the 1990 review, this alternative had the least impact on TVA’s
                           systemwide cost of supplying electric power. All of the alternatives
                           reduced TVA’s ability to generate hydroelectric power during periods of
                           peak power demand. TVA estimated that the selected alternative would
                           require replacing some hydroelectric power with power from more
                           expensive generation sources, at an annual average cost of $2 million (in
                           1990 dollars).

                           Since the 1990 review, little has changed in how TVA operates its
                           multipurpose tributary projects. Because it had been receiving an
                           increasing number of requests to analyze changes in the lake levels for
                           individual lakes, TVA determined that a piecemeal approach raised
                           questions of fairness in how each lake would be treated within TVA’s
                           system. TVA officials also believed that there was great uncertainty
                           surrounding the future of the electric utility industry due to deregulation
                           and restructuring. Therefore, in March 1997, TVA established a 4-year
                           moratorium on making any changes in lake levels.

                           TVA has recognized that any future changes to its policies impacting lake
                           levels require further study. In July 1998, an internal TVA task force report
                           recommended that TVA continue its moratorium and start reevaluating
                           policies impacting lake levels within the next 2 to 4 years. The task force
                           also noted the complexities involved in carrying out such a study and
                           identified several areas requiring further attention. Among these are a
                           proactive communication plan with the public and better evaluation
                           methodologies for costs and benefits. GAO agrees that further study is
                           warranted and that public and other stakeholder involvement is critical to
                           the success of TVA’s reexamination efforts. Equally important, however, is
                           a comprehensive analysis of the costs and benefits of any alternatives
                           analyzed. This report makes a recommendation directed to the Chairman
                           of TVA’s Board of Directors to ensure these actions take place.



Principal Findings

TVA Balances Various       The multipurpose tributary projects serve various purposes. These
Purposes of Multipurpose   purposes include maintaining a navigable waterway between Knoxville,
Tributary Projects         Tennessee, and Paducah, Kentucky; reducing the risk of flooding
                           throughout the Tennessee Valley, especially at Chattanooga, Tennessee;




                           Page 4                       GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                            Executive Summary




                            and providing hydroelectric power. In addition to these primary purposes,
                            these projects can also serve recreational and water quality purposes.

                            In operating these projects, TVA faces a balancing act of how to maximize
                            the benefits of the available water to meet all of the purposes. TVA’s ability
                            to lower and raise the lake levels during the year is a key element in this
                            balancing act. For example, TVA lowers the lake level for Douglas—a
                            multipurpose tributary project—50 feet from 990 feet on August 1 to 940
                            feet above sea level on January 1. TVA does so to reduce the risk of
                            flooding while also allowing for hydroelectric power production during
                            periods of peak power demand. When the flood risk diminishes, TVA allows
                            the lake levels to rise so that desirable summer recreation levels are
                            achieved by Memorial Day.


Changes Resulting From      Resulting from its 1990 review,2 a key policy change TVA implemented in
1990 Review and Cost        1991 was a 2-month delay in the annual lake drawdown of the
Impacts Estimated in 1990   multipurpose tributary projects. This change meant that the drawdown of
                            the lake levels could begin on August 1—some 2 months later than TVA’s
and 1999                    previous policy of the lake-level drawdown starting on Memorial Day.
                            Thus, lake levels were allowed to remain higher during June and July
                            under this new policy.3 Although various factors influenced TVA’s decision,
                            the impact on TVA’s systemwide cost of supplying electric power resulting
                            from the 2-month delay in the annual drawdown was the major factor TVA
                            considered when examining the seven alternatives. Other factors TVA
                            considered were minimizing the impacts on the environment and flood
                            risk in the Valley and increasing recreational opportunities at the lakes.
                            TVA selected the alternative having the lowest impact on its systemwide
                            cost of supplying electric power.

                            TVA used a complex methodology in estimating this impact. TVA estimated
                            an annual average increase of $2 million (in 1990 dollars) in its systemwide
                            cost of supplying electric power for the lake-level alternative selected.
                            Other alternatives were not selected, including maintaining the lake levels
                            higher until Labor Day or October 31, because the annual average


                            2
                             Tennessee River and Reservoir System Operation and Planning Review, TVA (Dec. 1990). This review
                            examined a number of issues, including potential changes in the lake levels at the multipurpose
                            tributary projects.
                            3
                             This policy also allowed TVA to fill the affected lakes above the August 1 levels. By doing so, TVA
                            could conduct a limited drawdown, referred to as a “restricted” drawdown, of the lake levels during
                            June and July and use this additional water to generate hydroelectric power during periods of peak
                            power demand in these 2 months, while still meeting the lake levels on August 1, when “unrestricted”
                            drawdown could begin.



                            Page 5                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                            Executive Summary




                            increases in TVA’s systemwide cost of supplying electric power were
                            estimated at $84 million and $93 million, respectively (in 1990 dollars).
                            However, TVA’s methodology did not attempt to quantify other types of
                            costs, such as impacts to flood control and navigation operations. In
                            addition, TVA did not attempt to quantify potential benefits that may result
                            from increased recreation or tourism by maintaining summer lake levels
                            longer.

                            To illustrate what the potential future impacts could be on TVA’s
                            systemwide cost of supplying electric power given the considerable
                            changes in the electricity industry since 1990, GAO requested that TVA
                            analyze two alternatives that were similar to those TVA examined in the
                            1990 review. These two alternatives were judgmentally selected to show
                            the potential impact of an additional 1-month drawdown delay for all of
                            the projects and an additional 2-month drawdown delay for three of the
                            projects. In 1990, TVA adopted a policy to begin the drawdown of the lake
                            levels on August 1. It was this policy against which TVA evaluated the
                            alternatives in 1999. By contrast, in 1990 TVA evaluated alternatives against
                            a Memorial Day drawdown date. In the 1999 update, as in 1990, TVA
                            identified systemwide cost increases of supplying electric power.

                            TVA’s1999 analysis showed that delaying drawdown of the lake levels at
                            the multipurpose tributary projects from August 1 until Labor Day could
                            result in estimated increased systemwide costs of supplying electric power
                            ranging from $0 to $88 million annually, with an average annual estimated
                            cost of $47 million (in 1999 dollars). The results for the second alternative
                            examined—three of the lakes having an October 1 drawdown date with
                            others keeping the August 1 drawdown—showed potential cost impacts
                            ranging from a $2 million decrease in costs to a $33 million increase in
                            costs annually, with an average annual estimated cost increase of
                            $14 million (in 1999 dollars). TVA cautioned, however, that both the 1990
                            and 1999 estimates were subject to a great deal of uncertainty due to
                            future hydrological conditions, electricity prices, and other variables.
                            While a complete evaluation of the cost-estimation methodologies used by
                            TVA was beyond the scope of GAO’s work, the general approach used by TVA
                            in 1990 and 1999 appeared to be reasonable.


Little Action Taken to      Although TVA continued to receive requests from individuals and
Change Project Operations   organizations during the 1990s to make additional changes to lake levels,
Since 1990 Despite          its policies have changed little. In March 1997, TVA adopted a 4-year
                            moratorium on making any changes to lake levels because it believed such
Requests From Users


                            Page 6                       GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                             Executive Summary




                             action would (1) position TVA better for future competition in the electric
                             utility industry, (2) minimize the public’s perception of favoritism for any
                             particular lake within the system, and (3) allow time for TVA to determine
                             how studies of lake levels should be evaluated in the future.

                             Even though TVA had implemented a moratorium, it commented on two
                             studies conducted by non-TVA organizations on the estimated benefits to
                             local economies resulting from lake levels being kept higher during the
                             year.4 TVA criticized certain aspects of the studies, such as the scope,
                             methodology, and assumptions. TVA noted that one of the studies lacked
                             proper recognition of the multipurpose roles served by TVA’s projects and
                             how changes would impact the entire system of projects. GAO also noted
                             that these studies were limited because they considered only the impacts
                             on TVA’s systemwide cost of supplying electric power that were estimated
                             by TVA and estimated benefits pertaining to only a few counties adjacent to
                             the lakes in question. Neither study performed a comprehensive analysis
                             of benefits and costs of proposed lake-level changes. Such comprehensive
                             analyses must be performed in order to provide a balanced evaluation.


TVA Plans Further Study of   TVA recognizes that further study of its policies impacting lake levels is
Multipurpose Tributary       warranted and that it must do more to prepare for an eventual
Projects                     reexamination. TVA’s internal task force has recommended that TVA
                             proceed slowly, however, because it needs to develop the evaluation
                             methods necessary to adequately perform a reexamination. The task force
                             reported that TVA needs to (1) better define flood risk impacts, (2) refine
                             water resource planning and operation models to better simulate the
                             operation of the integrated system of projects under various scenarios,
                             and (3) develop economic growth and development analysis methods to
                             better represent expected impacts under alternative policies impacting
                             lake levels. In addition, TVA does not want to set unrealistic expectations
                             about how quickly any decisions could be reached about whether changes
                             are or are not needed to policies impacting lake levels because of the
                             extensive time that may be required to examine the range of
                             environmental issues involved.

                             GAO agrees that a reexamination of TVA policies impacting lake levels is
                             warranted. GAO also agrees that formal and continuing communication
                             with the public and other stakeholders will be an extremely important

                             4
                               Economic and Fiscal Consequences of TVA’s Draw-Down of Cherokee and Douglas Lakes, Center for
                             Business and Economic Research, University of Tennessee (Oct. 1998) and The Economic Impact of
                             Alternate TVA Lake Management Policies, North American Water Management Institute, Inc., Athens,
                             GA (Dec. 1997).



                             Page 7                              GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                  Executive Summary




                  factor in TVA’s reexamination. These communications are needed to
                  (1) further educate TVA regarding the concerns and needs of the various
                  stakeholders that must be considered in the reexamination process,
                  (2) give TVA additional opportunities to explain the operation of its
                  integrated system and the complexities involved in evaluating changes to
                  the system, (3) establish realistic expectations of the time required to
                  reevaluate changes in policies impacting lake levels, (4) keep the public
                  informed of TVA’s ongoing activities and progress achieved, and
                  (5) increase the overall credibility of the reexamination process.

                  Past evaluations examining changes to lake levels have tended to
                  emphasize either the costs associated with the potential change as has
                  been the case with TVA’s efforts or localized economic benefits as has been
                  the case with studies performed by non-TVA organizations. When
                  reexamining any potential changes in lake levels, a balanced and
                  comprehensive decision can only be reached through consideration of the
                  costs and benefits of the alternatives examined.


                  GAO recommends that the Chairman of TVA’s Board of Directors (1) provide
Recommendation    for a formal and continuing communication process for the public and
                  other stakeholders to actively participate in TVA’s efforts to reexamine its
                  policies impacting lake levels and (2) ensure that TVA’s reexamination
                  efforts consider the costs and benefits of any potential changes to policies
                  impacting lake levels.


                  GAO  provided a copy of a draft of this report to TVA for its review and
Agency Comments   comment. TVA’s comments and GAO’s responses to those comments are
                  included as appendix I. TVA also provided some technical clarifications
                  that have been incorporated in this report where appropriate. TVA stated
                  that GAO had conducted a comprehensive assessment of TVA’s tributary
                  lake operating policies and that the draft report fairly summarized the
                  issues influencing TVA’s operations. Regarding GAO’s recommendation that
                  TVA provide for a formal and continuing communication process, TVA
                  stated that it is essential that the public continue to be involved in
                  decisions that affect how the Valley’s water resources are used. TVA also
                  recognized that there is an opportunity to improve its communications
                  with stakeholders. TVA added that it remains committed to communicating
                  fully with its stakeholders and others who depend on the integrated
                  management of the Tennessee River system.




                  Page 8                      GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Executive Summary




Regarding GAO’s recommendation that TVA’s reexamination efforts consider
the costs and benefits of potential changes, TVA stated that in operating the
river system for the greatest public benefit, it continues to look for new
methodologies that will provide for a more objective analysis of the
tradeoffs among competing demands. TVA stated that certain water uses,
such as hydropower and flood control, lend themselves more readily to
quantitative analysis, while other operating objectives, such as economic
development and environmental impacts, continue to be more difficult to
quantify. TVA added that it remains committed to providing lake users and
other beneficiaries with the best information it has on the most likely
impacts of changes in lake operations so that such lake users and other
beneficiaries are aware of the tradeoffs and consequences of policy
changes.

TVA also emphasized in its comments that not all costs are monetary. TVA
stated that as it works to meet multiple needs with a finite resource,
increased costs can take the form not only of higher electricity prices but
also reduced flood control benefits, lessened environmental quality, and
reductions in other benefits. TVA added that costs in any form become
costs to some segment of the public.




Page 9                       GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Contents



Executive Summary                                                                                    2


Chapter 1                                                                                           14
                        Background                                                                  14
Introduction            Objectives, Scope, and Methodology                                          17


Chapter 2                                                                                           22
                        TVA’s 54 Projects Exist for a Variety of Reasons                            22
Multipurpose            Navigation, Flood Control, and Hydropower Are Priority                      30
Tributary                 Purposes for Multipurpose Tributary Projects
                        Several Factors Influence How TVA Operates Its Multipurpose                 33
Projects—Why Do           Tributary Projects
They Exist and How      TVA Faces Balancing Act Between Operating Priorities and                    36
Are They Operated         Users’ Interests


Chapter 3                                                                                           39
                        1971 Study Led to Higher Winter Lake Levels                                 39
Multipurpose            1990 Review Resulted in Latest Changes in TVA’s Policies                    40
Tributary Project         Impacting Lake Levels
                        1990 Review Results in Delay of Annual Lake-Level Drawdown                  41
Operations Were           Date Until August 1
Changed as a Result     TVA’s 1990 Review and Its 1999 Analysis of Two Lake-Level                   49
of TVA’s 1990 Review      Alternatives Both Estimate Increases in TVA’s Systemwide Cost
                          of Supplying Electric Power
                        Conclusions                                                                 50


Chapter 4                                                                                           52
                        In 1997 TVA Decided to Establish 4-Year Moratorium on Changes               52
Little Has Changed in      to Project Operations
Multipurpose            December 1997 Request for Changes to Policies Impacting Lake                55
                           Levels From Users of Douglas Lake
Tributary Project       Study of the Economic and Fiscal Consequences of TVA’s                      56
Operations Since the       Draw-Down of the Cherokee and Douglas Lakes
1990 Review Despite     Mountain Lakes Study Committee Report on Georgia Lakes                      56
                        Conclusions                                                                 57
Requests From Users




                        Page 10                   GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                       Contents




Chapter 5                                                                                           58
                       TVA Recognizes That Various Lake-Level Issues Need to Be                     58
Formal and               Addressed
Continuing             TVA Has Several Options Available When Assessing the                         63
                         Environmental Effects of Alternative Policies Impacting Lake
Communication            Levels
Process With the       Conclusions                                                                  64
Public and Other       Recommendation                                                               65
                       Agency Comments and Our Evaluation                                           65
Stakeholders Needed
in Future Lake-Level
Study
Appendixes             Appendix I: Comments From the Tennessee Valley Authority                     68
                       Appendix II: Multipurpose Tributary Lake-Level                               70
                         Changes—Minimum and Maximum Monthly Levels During
                         Calendar Year 1998
                       Appendix III: TVA’s 1990 Review and Its 1999 Analysis                        73
                       Appendix IV: Study Examining Cherokee and Douglas Lakes                      83
                       Appendix V: Study Examining Georgia Mountain Lakes                           88
                       Appendix VI: Major Contributors to This Report                               92


Tables                 Table 2.1: August 1 and January 1 Target Lake Levels for TVA’s               26
                         Multipurpose Tributary Projects
                       Table 2.2: Various Characteristics of TVA’s Multipurpose                     36
                         Tributary Lakes
                       Table 3.1: Changes Made in 1971 to Multipurpose Tributary Lake               40
                         Levels
                       Table 3.2: Lake-Level Alternatives Analyzed by TVA During 1990               42
                         Review
                       Table 3.3: Effect of 1990 Changes to Multipurpose Tributary                  45
                         Median Lake Levels
                       Table 3.4: Multipurpose Tributary Lake Levels—Minimum and                    46
                         Maximum Level in Calendar Year 1998
                       Table 3.5: Multipurpose Tributary Lake Level Targets Throughout              47
                         the Year
                       Table 4.1: Comparison of Douglas Lake Monthly Elevation                      55
                         Differences—LOUD’s Alternative vs. TVA’s Median Lake Levels
                       Table 5.1: Funding Budgeted for TVA Activities Related to the                60
                         Lake Level Policy Task Force’s Recommendations




                       Page 11                    GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
          Contents




          Table III.1: Estimated Impact of Lake-Level Alternatives                      75
            Evaluated by TVA in 1990 on TVA’s Systemwide Cost of
            Supplying Electric Power
          Table III.2: Estimated Impact of Lake-Level Alternatives                      78
            Evaluated by TVA in 1999 on TVA’s Systemwide Cost of
            Supplying Electric Power—Differences Between Alternatives and
            Base Case
          Table IV.1: Comparison of Measures of Economic Impacts                        85
            Resulting From Higher Lake Levels in August and September for
            the Douglas and Cherokee Lakes
          Table V.1: Incremental Impacts of Lake Visitation for Three                   89
            Georgia Mountain Lakes


Figures   Figure 1.1: TVA Service Area                                                  17
          Figure 2.1: Location of TVA’s Multipurpose Tributary Projects                 24
            and Multipurpose Main River Projects
          Figure 2.2: How TVA’s Multipurpose Tributary Projects and                     25
            Multipurpose Main River Projects Are Integrated Into the
            Tennessee River
          Figure 2.3: Fontana Multipurpose Tributary Project Prior to and               28
            After Drawdown
          Figure 2.4: Hiwassee Multipurpose Tributary Project Prior to and              29
            After Drawdown
          Figure 2.5: Monthly Rainfall, Runoff and Flood Storage Allocation             35
            Above Chattanooga




          Abbreviations

          CLUA       Cherokee Lake Users Association
          EA         environmental assessment
          EIS        environmental impact statement
          GAO        General Accounting Office
          HLOLE      Hourly Loss of Load Expectation
          LOUD       Lake Owners and Users of Douglas
          MW         megawatt
          NEPA       National Environmental Policy Act
          TVA        Tennessee Valley Authority
          USFS       U.S. Forest Service
          WSM        Weekly Scheduling Model


          Page 12                     GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Page 13   GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 1

Introduction


               The Tennessee Valley Authority (TVA), a wholly owned government
               corporation, is, by some measures, the nation’s largest electric power
               producer. TVA, as a multipurpose, independent federal corporation, is
               responsible for managing both power and nonpower programs,1 including
               the use, conservation, and development of the natural resources related to
               the Tennessee River. In carrying out these responsibilities, TVA relies on an
               integrated system of 54 dams and associated components—referred to as
               projects—to harness the Tennessee River and its tributaries to serve
               various purposes for the public’s benefit. These purposes include
               maintaining navigable waterways, protecting the public from floods,
               producing hydroelectric power, and providing recreational opportunities
               on TVA lakes. Concerns have been raised about TVA’s operation of a key
               component of its integrated system—the multipurpose tributary projects,
               consisting of dams and lakes on the tributaries to the Tennessee River,
               such as the Cherokee and Douglas projects—to satisfy the multiple and
               often conflicting public benefits.


               TVA was established by the Tennessee Valley Authority Act of 1933 to,
Background     among other things, improve the navigability and to provide for the flood
               control of the Tennessee River and provide for the agricultural and
               industrial development of the Tennessee Valley. The generation and
               transmission of hydroelectric power was established by law as an
               additional benefit resulting from these activities, secondary only to river
               operations to support navigation and flood control. Over time, other
               project purposes beyond navigation, flood control, and hydroelectric
               power production began to be recognized for various projects. Two such
               purposes were recreation and water quality. These purposes, however, are
               subordinate to the statutory purposes of navigation, flood control, and
               hydropower.

               To carry out its river management responsibilities, TVA constructed or
               acquired 54 projects primarily along the Tennessee River and its
               tributaries as an integrated system to serve many purposes.2 These 54
               projects have a dam and can include several components, such as a lake
               behind the dam and hydroelectric power facilities at the dam. Some of
               these projects were constructed for specific purposes, such as producing
               hydroelectric power. Other projects serve multiple purposes. For example,
               13 of the projects located on the tributaries of the Tennessee River are

               1
                 See Tennessee Valley Authority: Information on Nonpower Programs (GAO/RCED-98-133R, Mar. 31,
               1998) for additional information.
               2
                One of the 54 projects is located on a tributary of the Cumberland River.



               Page 14                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 1
Introduction




classified as multipurpose tributary projects serving various purposes,
such as navigation, flood control, hydroelectric power production, and/or
recreation. In addition, there are other multipurpose projects located on
the main portion of the Tennessee River, and these projects serve, among
other things, navigation and power production purposes, but have limited
flood control capabilities.

TVA  has guidelines, which it refers to as lake-level policies, that prescribe
the levels that must be maintained at the lakes at various times during the
year. For example, TVA has established certain lake levels—referred to as
“target” levels—that should be met on January 1, March 15, June 1, and
August 1. The January 1 and March 15 targets are maximum levels and the
June 1 and August 1 targets are minimum levels. Under these guidelines,
most of the multipurpose tributary projects are subject to significant
lake-level changes during the year as a result of TVA’s flood control efforts.
Annually, TVA lowers the lake levels at these projects during the late
summer and fall in order to provide additional storage space so that the
projects can help control flooding. This storage space, which TVA refers to
as flood control capacity,3 is designed into the multipurpose tributary
projects4 so that rainfall and runoff in the winter and early spring can be
held back behind the dams to help ease or potentially avert a flooding
situation downstream from the dam. According to TVA, while large storms
can occur throughout the year, the major regional floods on the Tennessee
River normally occur between December and April. This lowering of water
levels for flood control purposes, a process referred to as the
“drawdown” of the lakes, occurs between August 1 and January 1, and
also allows TVA to generate hydroelectric power with the water released
from the dams during periods of peak power demand in the summer.
During the winter, the water levels are allowed to increase slowly through
mid-March from rainfall and runoff flowing into the lakes, and then more
rapidly through Memorial Day, allowing for increased recreational uses of
the lakes during the late spring/early summer period.

Over the years, TVA has performed studies and reviews examining whether
changes in lake levels could be made to improve recreational uses of TVA


3
 For example, three of the multipurpose tributary projects have over 3.7 million acre-feet of flood
control capacity available on January 1. An acre-foot is a unit of measurement used to describe the
amount of storage in a reservoir. One acre-foot equals the volume of water covering 1 acre (43,560
square feet) to a depth of 1 foot.
4
 For this report, we define multipurpose tributary projects as including 14 projects—the 13 projects
mentioned (Boone, Chatuge, Cherokee, Douglas, Fontana, Hiwassee, Melton Hill, Norris, Nottely,
South Holston, Tellico, Tims Ford, and Watauga) and another project (Blue Ridge) because this project
has an annual drawdown cycle similar to the other 13 projects.



Page 15                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 1
Introduction




lakes. The most recent review was published in December 1990. This 1990
review entitled, Tennessee River and Reservoir System Operation and
Planning Review, resulted in TVA delaying the annual lake-level drawdown
at multipurpose tributary projects from Memorial Day to August 1. This
delayed lake-level drawdown allowed lake levels to remain higher during
the summer recreation season.

TVA supplies the energy needs of about 8 million people over a service area
covering 80,000 square miles, including most of Tennessee and parts of six
surrounding states (see fig. 1.1). In fiscal year 1998, TVA had about 28,500
megawatts of generating capacity,5 which consisted of 11 fossil plants (59
units), 3 nuclear plants (5 units), 29 hydroelectric plants (with 109 units), 4
combustion turbine plants (48 units), and 1 pumped storage plant6 (with 4
units). TVA’s hydroelectric power facilities constitute slightly over 19
percent, or about 5,500 megawatts, of TVA’s total generating capacity. TVA
had total operating revenues of $6.7 billion in fiscal year 1998 and
generated about 155 billion kilowatt-hours of energy, with TVA’s
hydropower facilities accounting for about 15.7 billion kilowatt-hours, or
about 10 percent, of that energy.




5
 TVA’s total generating capacity includes 405 megawatts from the U.S. Army Corps of Engineers’
projects on the Cumberland River system.
6
 A pumped storage plant is designed to generate hydroelectric power during peak periods of demand
by releasing water previously pumped into an elevated storage reservoir, usually during periods of low
power demand.



Page 16                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                Chapter 1
                                Introduction




Figure 1.1: TVA Service Area




                                                                Kentucky
                                                                                            Virginia




                                                                                          North
                               Tennessee                                                  Carolina




                                                               Georgia


     Mississippi                 Alabama




                                Source: TVA.



                                Representative Van Hilleary requested that we provide him with
Objectives, Scope,              information showing how TVA manages and operates its multipurpose
and Methodology                 tributary projects for various purposes, such as flood control, navigation,
                                hydropower production, and recreation. Specifically, we provide
                                information on (1) the purposes served by TVA’s multipurpose tributary
                                projects and how TVA operates these projects within its integrated system,
                                (2) the operational changes TVA made to these projects as a result of its




                                Page 17                     GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 1
Introduction




December 1990 Tennessee River and Reservoir System Operation and
Planning Review and the major factors influencing these changes, (3) the
actions TVA has taken since the 1990 review to address requests for
changes in the way TVA operates these projects, and (4) TVA’s plans for any
future changes in the way it operates these projects. We also provide
information on a selected update by TVA of its analysis performed in the
1990 review.

The primary agency included in our work was TVA. We contacted other
agencies, including the U.S. Army Corps of Engineers, to obtain
information on the effects of TVA’s operations on various project purposes,
such as water quality or navigation. We also visited and/or contacted a
limited number of stakeholders affected by TVA’s project operations, such
as lake user groups, commercial businesses located on or downstream
from the projects, local and state government officials, and state and
federal environmental officials. The purpose of these discussions was to
gain a sense of some of the issues concerning TVA stakeholders. At each
entity, we interviewed officials and obtained pertinent records, as
appropriate.

To determine the purposes served by TVA’s multipurpose tributary projects
and how TVA operates its system of projects to satisfy these purposes, we
examined TVA documentation and authorizing legislation and interviewed
TVA officials. We obtained information describing TVA’s operation of its
multipurpose tributary projects and how TVA balances the various
purposes for which these projects were authorized. We also held
discussions with a limited number of stakeholders affected by TVA’s
operations in order to describe the various issues that TVA faces in the
operation of its projects and how the operation affects such stakeholders.
In addition, we toured two of the multipurpose tributary lakes—Cherokee
and Douglas—in the fall of 1998 to view the extent of drawdown that had
taken place since August 1, 1998.

In order to describe the operational changes TVA made in 1991 to its
multipurpose tributary projects as a result of its December 1990 Tennessee
River and Reservoir System Operation and Planning Review and what
major factors influenced those changes, we examined TVA’s 1990 review
and interviewed TVA officials. We also reviewed information on the
alternative drawdown dates that TVA examined, the constraints and issues
considered in TVA’s review, and the methodologies used to determine the
additional energy and capacity costs associated with delaying the
drawdown of the tributary projects. In addition, we requested that TVA



Page 18                     GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 1
Introduction




analyze two alternatives included in its 1990 review to provide a more
current estimate of the increase in TVA’s systemwide cost of supplying
electric power resulting from the implementation of these alternatives.
One of these alternatives involved extending the drawdown to start on
Labor Day for all of the projects. The other alternative involved extending
the drawdown to start on October 1 for three projects, with the other
projects maintaining an August 1 drawdown date. We judgmentally
selected these alternatives to show the potential impact of a 1-month
drawdown delay for all of the projects and an additional 2-month delay for
three of the projects.

For TVA’s 1990 estimates of the impacts on TVA’s systemwide cost of
supplying electric power resulting from lake-level alternatives, we
reviewed the December 1990 Tennessee River and Reservoir System
Operation and Planning Review. We also reviewed the available
documentation that TVA officials provided to us on their 1990 cost
estimates and discussed with them the data and methodology they used
for their estimates.

We also held discussions with TVA officials to determine to what extent
market conditions and evaluation methodologies have changed since 1990.
Because changes have occurred since 1990, we requested that TVA analyze
two of the alternatives evaluated in 1990 to illustrate what the potential
impacts of these alternatives could be in the future. TVA officials provided
us with oral presentations of the methodology used and the results of the
1999 update. In addition, the officials provided us with a summary of their
assumptions, methodology, caveats, and results. In conducting the 1999
analyses, TVA considered its planned and/or ongoing efforts to purchase
peaking power (that is, power needed for the periods of greatest power
demand) and install additional natural gas combined-cycle combustion
turbines.

We developed a general understanding of TVA’s cost-estimation
methodology, including the three computer models used in the cost
estimation. The three models are: (1) the Weekly Scheduling Model (WSM)
of TVA’s hydrological and hydroelectric system, (2) the PowrSym power
production costing model, and (3) the Hourly Loss of Load Expectation
(HLOLE) capacity planning model. Some of the data inputs that TVA used,
specifically its forecast of electricity market prices over the next 25 years,
are proprietary TVA data.




Page 19                      GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 1
Introduction




We asked TVA officials whether the data and methodology that it used to
estimate the impact on its systemwide cost of supplying electric power
were reviewed by internal or external reviewers. We also asked whether
TVA uses the same data and similar methodology for its own operations
and planning. A contractor tested the HLOLE model and found it to be
accurate for a large power system such as TVA’s. TVA reported that it tested
WSM prior to the 1990 cost analysis and that it performed well. TVA also
reported that it uses the same data (including the electricity price
forecast) and models for its own capital budgeting decisions and other
internal purposes. We did not independently evaluate the reliability of the
data and the models that TVA used.

To determine what actions TVA has taken since the 1990 review to address
requests for changes in the way the multipurpose tributary projects are
operated, we examined TVA’s policies since 1991 regarding how it
addresses requests for changes to its operations and interviewed TVA
officials. We examined information showing why and when such policies
were implemented. We also reviewed two recent studies (conducted by
groups external to TVA) showing estimates of economic benefits to local
communities due to changes to policies impacting lake levels associated
with TVA lake-level drawdown activities at certain lakes and TVA’s
responses to these studies. We examined the first study, Economic and
Fiscal Consequences of TVA’s Draw-Down of Cherokee and Douglas Lakes,7
 which was issued in October 1998, and met with the principal authors
from the University of Tennessee’s Center for Business and Economic
Research. We also examined the second study, The Economic Impact of
Alternate TVA Lake Management Policies,8 which concentrated on
economic impacts of TVA’s policies on three north Georgia lakes—Blue
Ridge, Chatuge, and Nottely. We met with officials from the North
American Water Management Institute, Inc., who prepared the study for
interested stakeholders in north Georgia. We also discussed the strengths
and weaknesses of both these recent studies with the authors and with TVA
officials. We also evaluated these studies using water project evaluation
guidelines of the U.S. Water Resources Council. While it is not required for
privately commissioned studies to follow these guidelines, we used them
because, in our view, they constitute the best available guidance on this
type of economic analysis.



7
Prepared for Land Owners and Users of Douglas by the Center for Business and Economic Research,
University of Tennessee, (Oct. 1998).
8
 Prepared for the Mountain Lakes Study Committee by the North American Water Management
Institute, Inc., Athens, GA, (Dec. 1997).



Page 20                             GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 1
Introduction




To explain TVA’s plans for any future changes in the way it operates the
multipurpose tributary projects, we held discussions with TVA officials and
examined TVA documentation to determine TVA’s current efforts and
potential plans for any future review of its operations. In addition, we
reviewed budgetary documentation regarding activities that TVA has
funded for fiscal year 1999 and/or budgeted for future years for the
examination of any new or improved analytical tools aimed at evaluating
how changes in lake levels affect flood risk, navigation, and economic
development.

We also obtained TVA’s views on the necessity of an environmental impact
statement with regards to any future changes to TVA’s operating policies
affecting the multipurpose tributary projects. We reviewed the
requirements of the National Environmental Policy Act of 1969 (NEPA) and
pertinent court cases from recent years to determine how the NEPA
requirements might apply to TVA and its policies impacting lake levels.

We did not independently verify the data we obtained from TVA or the
other entities we contacted.

We conducted our review from October 1998 through May 1999 in
accordance with generally accepted government auditing standards.




Page 21                     GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 2

Multipurpose Tributary Projects—Why Do
They Exist and How Are They Operated

                         TVA’s network of projects exists for a variety of reasons and is operated for
                         many purposes. In operating these projects, TVA is guided by the operating
                         priorities contained in the TVA Act. These priorities require that TVA operate
                         its system of projects primarily to promote navigation and flood control
                         and, to the extent consistent with these purposes, for hydroelectric power
                         production. While TVA uses two sets of broadly defined policies or
                         guidelines—one for lake levels and the other for reservoir releases—to
                         guide the operation of the projects within the integrated system,
                         operational limitations exist at these projects due to the project design
                         characteristics. Although some of the multiple purposes served by these
                         projects conflict, TVA attempts to maximize the benefits of the Tennessee
                         River while adhering to the operating priorities of the projects.


                         TVA has constructed or acquired 54 projects which differ in age, size, and
TVA’s 54 Projects        authorized purposes. One type of these projects—the multipurpose
Exist for a Variety of   tributary project—consists of dams and lakes on the tributaries to the
Reasons                  Tennessee River and provides multiple public benefits. These projects
                         have significant changes in lake levels during the year to support flood
                         control and hydroelectric power production efforts.

                         TVA  defines its projects in four categories: multipurpose tributary projects,
                         multipurpose main river projects, single-purpose power projects,1 and
                         nonpower tributary projects. Figure 2.1 shows the location of the
                         multipurpose tributary projects and the multipurpose main river projects
                         within the Tennessee Valley. Figure 2.2 illustrates how these projects are
                         integrated into the Tennessee River. These figures were provided to us by
                         TVA.



Multipurpose Tributary   TVA has 13 multipurpose tributary projects, located on various tributaries
Projects                 connecting to the Tennessee River. These projects were constructed to
                         serve multiple purposes, including hydroelectric power production and
                         one or more of the following: flood control, navigation, recreation, and
                         water supply. Most of the projects in this category have a significant
                         amount of available storage space for flood control purposes. As a result,
                         TVA operates these projects primarily for flood control and hydroelectric
                         power production purposes and annually draws the lake levels down in

                         1
                          For this report, we have included one of these projects—Blue Ridge—in our definition of a
                         multipurpose tributary project because (1) even though it is a single-purpose power project, it has an
                         annual lake-level drawdown similar to most of the 13 multipurpose tributary projects and (2) TVA has
                         considered it as being very similar to the multipurpose projects in past reviews involving such
                         projects.



                         Page 22                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                           Chapter 2
                           Multipurpose Tributary Projects—Why Do
                           They Exist and How Are They Operated




                           the late summer and fall to help ease or potentially avert flooding in the
                           winter. The multipurpose tributary projects are: Boone, Chatuge,
                           Cherokee, Douglas, Fontana, Hiwassee, Melton Hill, Norris, Nottely, South
                           Holston, Tellico,2 Tims Ford, and Watauga.


Multipurpose Main River    Along the main portion of the Tennessee River, TVA operates and maintains
Projects                   nine multipurpose main river projects, which serve multiple purposes,
                           including navigation and hydroelectric power production and, in most
                           cases, flood control. In comparison to the multipurpose tributary projects,
                           these projects maintain fairly stable lake levels. However, most of the
                           projects in this category have some available flood storage space for water
                           in the lake, and as a result, TVA operates these projects for flood control
                           purposes and conducts a limited drawdown of the lake levels to prepare
                           for potential flooding events in the winter.


Single-Purpose Tributary   TVA maintains and operates 10 single-purpose tributary projects on
Projects                   tributaries of the Tennessee River that were constructed or acquired
                           strictly for hydroelectric power production.3 However, these projects also
                           benefit other purposes, including recreation, water supply and, in the
                           instance of the Blue Ridge project, flood control. Most of these projects
                           have no storage space available in the lake for flood control purposes, and
                           the water that flows to the project is generally sent directly through the
                           hydroelectric power facilities to generate power.


Nonpower Tributary         TVA has 22 projects located on tributaries to the Tennessee River that do
Projects                   not have hydroelectric power facilities and are operated for a variety of
                           purposes, such as flood control, recreation, and water supply. Some of
                           these projects have storage space within the lakes, and TVA operates these
                           projects for flood control purposes and draws the lake levels down to
                           prepare for potential flooding events in the winter. Some of the projects
                           are essentially overflow structures, requiring little operator intervention
                           except during special maintenance operations.




                           2
                           This project does not have a power facility; the water from it is diverted to the Fort Loudoun project,
                           which has a power facility.
                           3
                            One project is located on the Caney Fork River, a tributary of the Cumberland River. For this report,
                           the project is classified as a Tennessee River single-purpose tributary project. In addition, another
                           project has no hydroelectric facilities, but it provides a small reservoir for cooling water intake at a
                           fossil plant.



                           Page 23                                 GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                                   Chapter 2
                                                   Multipurpose Tributary Projects—Why Do
                                                   They Exist and How Are They Operated




Figure 2.1: Location of TVA’s Multipurpose Tributary Projects and Multipurpose Main River Projects




                                Kentucky




                                                                                                                                        Boone         South Holston

                                                                                                    Norris                                  Watauga
                                                                                                                             Cherokee

                                                                                              Melton Hill
                                                                                                                            Douglas
                                                                                                              Ft. Loudoun
                                                                                                Tellico
                                                                                  Watts Bar
                                                                                                                    Fontana

                                                             Tims Ford   Chickamauga                        Hiwassee
                               Pickwick
                                                                                                                         Chatuge
                                                                           Nickajack
                                               Wheeler                                                           Nottely

                                          Wilson
                                                                                                            Blue Ridge
                                                         Guntersville




     Dam

  Multipurpose Tributary Projects           Multipurpose Main River Projects
  Blue Ridge                                Chickamauga
  Boone                                     Fort Loudoun
  Chatuge                                   Guntersville
  Cherokee                                  Kentucky
  Douglas                                   Nickajack
  Fontana                                   Pickwick
  Hiwassee                                  Watts Bar
  Melton Hill                               Wheeler
  Norris                                    Wilson
  Nottely
  South Holston
  Tellico
  Tims Ford
  Watauga




                                                   Page 24                              GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                          Chapter 2
                                          Multipurpose Tributary Projects—Why Do
                                          They Exist and How Are They Operated




Figure 2.2: How TVA’s Multipurpose
Tributary Projects and Multipurpose
Main River Projects Are Integrated Into
the Tennessee River




Multipurpose Tributary                    While all 54 projects were built or acquired as part of TVA’s integrated
Projects Have Significant                 system of projects and all of the projects contribute to TVA’s attempts to
Lake-Level Changes                        maximize the value of the available water in the Tennessee River, the



                                          Page 25                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                    Chapter 2
                                    Multipurpose Tributary Projects—Why Do
                                    They Exist and How Are They Operated




                                    multipurpose tributary projects generally have more significant changes in
                                    lake levels during the year. For example, the target lake level for
                                    Douglas—a multipurpose tributary project—decreases 50 feet from 990
                                    feet on August 1 to 940 feet above sea level on January 1. On the other
                                    hand, the target lake level for Fort Loudoun—a multipurpose main river
                                    project—only decreases 6 feet from 813 feet on August 1 to 807 feet above
                                    sea level on January 1. Table 2.1 shows differences between the August 1
                                    and January 1 target lake levels at the multipurpose tributary projects.

Table 2.1: August 1 and January 1
Target Lake Levels for TVA’s        In feet above sea level
Multipurpose Tributary Projects     Multipurpose tributary             August 1 target        January 1 target
                                    projects                                     level                   level               Difference
                                    Blue Ridge                                      1,682                   1,668                         14
                                    Boone                                           1,382                   1,358                         24
                                    Chatuge                                         1,923                   1,912                         11
                                    Cherokee                                        1,060                   1,030                         30
                                    Douglas                                           990                     940                         50
                                    Fontana                                         1,693                   1,644                         49
                                    Hiwassee                                        1,515                   1,465                         50
                                                                                          a                       a                        b
                                    Melton Hill
                                    Norris                                          1,010                     985                         25
                                    Nottely                                         1,770                   1,745                         25
                                    South Holston                                   1,721                   1,702                         19
                                    Tellico                                           813                     807                          6
                                                                                          c
                                    Tims Ford                                                                 873c                         b

                                    Watauga                                         1,949                   1,940                          9
                                    a
                                     These dates do not have a set level due to the small change in lake levels; the full range is from
                                    790 to 796 feet above sea level, with a normal operating range of 793 to 795 feet above sea level.
                                    b
                                        Not applicable.
                                    c
                                     The Tims Ford project has no August 1 target level. However, it does have a minimum elevation
                                    requirement of 883 feet above sea level from May 15 through October 15 and a January 1 target
                                    level of 873 feet above sea level.

                                    Source: TVA.




                                    Page 26                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 2
Multipurpose Tributary Projects—Why Do
They Exist and How Are They Operated




The changes in lake levels resulting from the annual drawdown at most of
these multipurpose tributary projects are significantly greater than the
lake-level changes in the multipurpose main river projects due to the
topography, the design of the projects, and the purposes for which they
were intended. Figures 2.3 and 2.4 show photographs, which TVA provided
to us, of what two multipurpose tributary projects look like as a result of
the annual lake-level drawdown between the summer and January 1.




Page 27                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                       Chapter 2
                                       Multipurpose Tributary Projects—Why Do
                                       They Exist and How Are They Operated




Figure 2.3: Fontana Multipurpose
Tributary Project Prior to and After
Drawdown




                                       Page 28                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                       Chapter 2
                                       Multipurpose Tributary Projects—Why Do
                                       They Exist and How Are They Operated




Figure 2.4: Hiwassee Multipurpose
Tributary Project Prior to and After
Drawdown




                                       Page 29                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                            Chapter 2
                            Multipurpose Tributary Projects—Why Do
                            They Exist and How Are They Operated




                            TVA primarily operates its multipurpose tributary projects for navigation,
Navigation, Flood           flood control, and hydroelectric power production, as specified in the TVA
Control, and                Act. Over time, however, TVA has recognized a variety of other purposes
Hydropower Are              that also benefit from TVA’s operation of its projects. These additional
                            purposes include recreation and water quality.
Priority Purposes for
Multipurpose            •   Navigation - TVA’s system of nine navigation locks at its main river projects
                            is part of a 650-mile transportation route from Knoxville, Tennessee, to
Tributary Projects          Paducah, Kentucky, which links the Tennessee Valley with the Mississippi
                            and Ohio River systems, connecting ports in 21 states. Commerce moved
                            on the Tennessee River has increased from about 2 million tons in the
                            1930s to about 49 million tons in 1997. According to TVA, over $450 million
                            worth of products are transported annually via barges through the river.
                            As a standard in the U.S. inland waterway industry that dates back to the
                            1920s, barges are designed for a 9-foot draft for navigation purposes.4 In
                            order to provide navigability on the Tennessee River, TVA maintains a
                            minimum channel depth of 11 feet. Under normal weather conditions, the
                            operation of the multipurpose tributary projects for flood control and
                            hydroelectric power production provides enough reliable water
                            downstream to maintain navigation depths on the river. However, during
                            dry periods, TVA may have to release water from certain main river and/or
                            tributary projects to meet required navigation depths. In addition, during
                            floods and low-flow periods, TVA’s integrated system affects navigation on
                            the Mississippi River and the lower Ohio River. For example, TVA may at
                            times, release water to help maintain navigable channels on the Ohio and
                            Mississippi Rivers during low-flow periods. TVA can also reduce water
                            releases from its projects to mitigate flooding on these rivers during
                            hazardous, high-flow periods.
                        •   Flood control - One of the primary purposes for TVA’s development and
                            maintenance of its integrated system of projects is to protect the
                            inhabitants of the Tennessee Valley from loss of property and life as a
                            result of flooding. An area within the Tennessee Valley that is very
                            susceptible to flooding is the city of Chattanooga, which sits at a point just
                            before the Tennessee River passes through the Cumberland Mountains.
                            Before the construction of TVA dams to control the flow of water, major
                            storms would force the river to flood its banks. While TVA’s system of
                            projects is not sufficient to eliminate all flooding, in most cases, it
                            significantly reduces the risk of major damages and loss of life. As of
                            April 1998, TVA estimated that the cumulative value of flood damages



                            4
                             The term “draft” refers to the vertical distance that towboats and barges extend below the water
                            surface.



                            Page 30                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
    Chapter 2
    Multipurpose Tributary Projects—Why Do
    They Exist and How Are They Operated




    prevented at Chattanooga by the operation of the TVA’s integrated system
    of projects since 1936 totaled approximately $4.4 billion.
•   Hydroelectric Power Production - TVA’s hydroelectric power facilities
    serve as its most economical and versatile source of electricity. Electricity
    generated by TVA’s hydroelectric power facilities, which are made up of
    113 hydropower units at 30 projects, constitutes approximately 10 percent
    of TVA’s annual total electricity production.5 While the lake levels are
    managed within a specific operating range, TVA attempts to produce as
    much electricity as possible from the available water at the times when the
    power is most valuable, without disrupting flood control and navigation
    requirements. Hydroelectric power production is considered the most
    economical source of electricity because its incremental costs of
    production are cheaper than any other power sources. In addition, TVA’s
    hydroelectric power facilities are more versatile than other power sources,
    due to the ability to turn units on more quickly than other units to meet
    peak power demands, system emergencies, and voltage regulation on the
    transmission system.

    TVA attempts to schedule releases of water through its hydroelectric power
    facilities at the tributary projects to avoid releasing more water into the
    Tennessee River than the main river hydropower facilities can handle.
    While water can be released downstream through a project’s spillways
    whenever necessary, TVA tries to minimize such releases in favor of
    releasing water through the hydropower facilities to produce electricity.
    To maximize the value of its hydroelectric power, TVA schedules the
    production of hydroelectric power during the periods of greatest power
    demand to offset the use of other, more expensive, sources of power. Due
    to power usage in the Valley, daily peak demands in the winter typically
    last a few hours in the morning and a few hours in the evening. Often,
    severely cold weather conditions in the winter only last for a few days at
    most. In contrast, the typical daily summer peak demands last 8 or more
    continuous hours a day and the extremely hot weather driving these
    demands can last multiple days or even weeks. Based on TVA’s demand
    forecasts, TVA officials predict that the summer peak demand for power
    will consistently exceed the winter peak demand in the future. While TVA
    schedules most of the hydroelectric power generation to help meet the
    daily peak power demands, at times hydroelectric power facilities are
    operated 24 hours a day while water is released for other purposes, such
    as water quality or flood control.



    5
     TVA’s hydroelectric power facilities include the Raccoon Mountain pumped storage project, which
    has four hydropower units.



    Page 31                              GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
    Chapter 2
    Multipurpose Tributary Projects—Why Do
    They Exist and How Are They Operated




    In addition to operating its multipurpose tributary projects for the
    statutory purposes of navigation, flood control, and hydropower
    generation, TVA also manages these projects for other purposes, if
    consistent with the three statutory purposes.

•   Recreation - TVA regulates its lakes to provide recreational opportunities
    consistent with its flood control, navigation, and power production
    responsibilities. TVA estimates that its lakes attract millions of visitors each
    year for a variety of water-related activities, including swimming, fishing,
    water skiing, and boating. TVA also provides recreational opportunities on
    the Tennessee River downstream from TVA projects. For example, TVA
    releases water at the Ocoee No. 2 project for whitewater recreation during
    certain times of the year.
•   Minimum flows for water quality - Prior to 1991, TVA had fewer
    requirements to provide minimum flows from its projects for the benefit of
    aquatic life downstream from TVA projects. While water quality benefits
    were often provided as a result of TVA operations for flood control, power
    production, and navigation, TVA could shut off water releases, with a few
    exceptions, downstream from its projects. Shutting down the flow of
    water downstream from TVA projects can have severe effects on aquatic
    life. In December 1990, TVA recommended that increased minimum flows
    be provided at some of its mainstream and tributary projects and dissolved
    oxygen levels be improved in water released downstream from 16 of its
    projects. TVA’s implementation of these recommendations helped recover
    over 180 miles of aquatic habitat and improved dissolved oxygen levels in
    over 300 miles of the Tennessee River. TVA also attempts to stabilize lake
    levels during the spring to support the spawning season for a variety of
    fish species that can be found in the Tennessee River and TVA’s lakes.
    These fish species deposit their eggs primarily during the spring, at various
    times from February through June, depending on the species and water
    temperature. In addition, minimum flow requirements affect other issues
    related to water quality, such as water supply, heated discharges from coal
    and nuclear plants, and wastewater discharges from industries.

    Water Supply – Water is withdrawn at approximately 330 points along the
    Tennessee River and its tributaries to benefit approximately 4 million
    citizens of the TVA region. TVA’s operation of its multipurpose tributary and
    mainstream projects provides the necessary water flows to allow water
    supply pumping mechanisms to function properly. On average, over
    9 billion gallons of water are withdrawn from the river system each day.
    According to TVA, over 75 percent of the water withdrawn is returned to a
    river, stream, or lake after use.



    Page 32                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                     Chapter 2
                     Multipurpose Tributary Projects—Why Do
                     They Exist and How Are They Operated




                     Heated Discharges from Coal and Nuclear Plants - TVA operates coal and
                     nuclear plants on or near lakes along the Tennessee River and its
                     tributaries for power-generation purposes. TVA provides minimum flows
                     from its multipurpose projects to supply cooling water for its coal and
                     nuclear power plants. In turn, the coal and nuclear plants discharge the
                     water back into the Tennessee River at a warmer temperature. If needed,
                     TVA provides releases from upstream multipurpose projects to reduce the
                     effects of the upstream flow of heated discharge water.

                     Wastewater Discharges from Industries - Industries operating along the
                     Tennessee River and its tributaries can apply for and receive approval for
                     permits from state pollution control agencies for the release of municipal
                     and industrial effluents into the Tennessee River and its tributaries. TVA
                     provides historical flow data to the state pollution control agencies to help
                     them set appropriate permit limits, based on TVA’s normal operations of its
                     projects. If an industry desires additional flow for the assimilation of its
                     effluent beyond what is provided from normal project operations, the
                     industry must reimburse TVA for the costs it incurs in providing these
                     flows.


                     In addition to the operating priorities contained in the TVA Act, there are
Several Factors      several factors that influence how TVA operates its multipurpose tributary
Influence How TVA    projects. TVA’s establishment of two sets of broadly defined policies or
Operates Its         guidelines—one for lake levels and the other for reservoir releases—is one
                     example. An additional constraint on operations is the specific design
Multipurpose         characteristics associated with each of the projects. Within the limitations
Tributary Projects   of such factors, TVA attempts to balance the various purposes served to
                     provide the greatest public benefits of the Tennessee River within the
                     parameters of its operating priorities and the individual project design
                     characteristics.

                     TVA follows two sets of broadly defined policies or guidelines in the
                     operation of its mainstream and tributary projects: lake-level policies and
                     reservoir release policies. Lake-level policies prescribe a maximum,
                     minimum, or range of lake levels that must be maintained at a given time
                     of year. TVA graphically presents the lake-level requirements on line graphs
                     called “operating curves,” which show the lake-level requirements
                     throughout the year for flood control purposes, target recreational levels,
                     and the range of water levels within which hydroelectric power can be
                     generated. Reservoir release policies prescribe a maximum or minimum




                     Page 33                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 2
Multipurpose Tributary Projects—Why Do
They Exist and How Are They Operated




flow that must be maintained from a project over an hourly, daily, weekly,
or biweekly period for navigation and water quality purposes.

On the tributary lakes, TVA’s policies require that the lake levels be lowered
annually during the summer through early winter to the projects’ January 1
target levels, primarily to help control potential floods during the winter
and early spring. Figure 2.5 (provided to us by TVA) shows that, based on
historical data of the TVA region, the wettest periods of the year tend to
occur between December and March. During this period, the days are
shorter, weather is colder, and the vegetation is dormant, which results in
a much greater amount of precipitation running off to the streams and
rivers than occurs in late spring, summer, and early fall. Thus, there is a
significantly higher risk of flooding during the January through March
period. The drawdown of the lake levels also provides for hydropower
generation to help meet the peak power demands of the summer and
provides augmented river flows for water quality and navigation purposes
during the summer and fall, historically dry periods of the year. As the
flood risk diminishes during late winter and early spring, TVA allows the
lakes to accumulate water in order to reach desirable summer levels by
Memorial Day. Prior to 1991, TVA began the annual drawdown of the lake
levels of its multipurpose tributary projects anytime after Memorial Day.
Resulting from the changes implemented in 1991, the lake-level drawdown
for these projects now begins August 1.




Page 34                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                          Chapter 2
                                          Multipurpose Tributary Projects—Why Do
                                          They Exist and How Are They Operated




Figure 2.5: Monthly Rainfall, Runoff and Flood Storage Allocation Above Chattanooga

6 Inches




5




4




3




2




1




0
                                                                                                                  er
                                                                 ly




                                                                               t
                   ry




                                                                                                      er
                                                                                            er
                                 ril




                                           y
                         rch




                                                                                                                            er
         ry




                                                                              us
                                                      e
                                         Ma




                                                               Ju




                                                                                                             mb
                                                       n
               ua




                                                                                                  tob
                                                                                          mb
                                Ap




                                                                                                                        mb
        a




                                                                               g
                                                    Ju
                        Ma
     nu




                                                                            Au
              br




                                                                                                             ve
                                                                                      pte




                                                                                                 Oc




                                                                                                                       ce
    Ja




              Fe




                                                                                                           No




                                                                                                                       De
                                                                                     Se




     Months
                                                Rainfall (100-year average)

                                                Runoff (100-year average)

                                                Flood storage allocation




                                          Another factor that TVA must consider in operating its integrated system of
                                          projects is the design characteristics of the projects. These characteristics
                                          are quite varied for the multipurpose tributary projects. For example, all,
                                          except one, have hydropower facilities, with installed generating capacity
                                          ranging from a low of 10 megawatts at Chatuge to a high of 239 megawatts
                                          at Fontana. All, except one, have the ability to alleviate potential flooding,
                                          with capacity at January 1 ranging from a low of 68,500 acre feet at Blue
                                          Ridge to three projects—Norris, Cherokee, and Douglas—each having over
                                          1 million acre feet. The total area of the lakes at normal summer maximum




                                          Page 35                                  GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                          Chapter 2
                                          Multipurpose Tributary Projects—Why Do
                                          They Exist and How Are They Operated




                                          ranged from 3,290 acres at Blue Ridge and 4,310 acres at Boone to 34,200
                                          acres at Norris. In addition, TVA manages shoreline miles at its projects,
                                          ranging from a low of 68 miles at Blue Ridge and 102 miles at Nottely to
                                          809 miles at Norris. One project—Melton Hill—has a navigation lock.
                                          Table 2.2 lists various characteristics for the multipurpose tributary
                                          projects.


Table 2.2: Various Characteristics of TVA’s Multipurpose Tributary Lakes
                                                                                                                     Area of
                                                                                                                     lake at   Installed
                                                                                                                     normal generating
                                                                                                     Length of      summer     capacity
                                                      Flood control capacity (in acre feet)          shoreline    maximum       (in total
Multipurpose tributary lake                              January 1      March 15          June 1     (in miles)   (in acres) megawatts)
Blue Ridge                                                  68,500         40,800          13,100           68        3,290           22
Boone                                                       92,400         60,400          12,900          127        4,310           81
Chatuge                                                     93,000         73,300          13,900          128        7,050           10
Cherokee                                                 1,011,800        807,800         146,700          395       30,300          135
Douglas                                                  1,251,000      1,008,600         237,500          512       30,400          146
Fontana                                                    580,000        580,000          73,400          238       10,640          239
Hiwassee                                                   270,200        216,100          35,000          165        6,090          166
                                                                    a               a            a
Melton Hill                                                                                                193        5,690           72
Norris                                                   1,472,800      1,113,000         512,000          809       34,200          131
Nottely                                                    100,000         79,100          12,300          102        4,180           15
South Holston                                              290,200        220,100         106,100          182        7,580           39
Tellico                                                    120,000        120,000          32,000          357       15,860        None
Tims Ford                                                  219,600        167,200          78,000          309       10,600           45
Watauga                                                    223,000        152,800         108,500          105        6,430           58
                                          a
                                           There is no flood storage allocation.



                                          Source: TVA.




                                          Many people and entities use the water in the Tennessee River and its
TVA Faces Balancing                       tributaries for various purposes and are directly affected by TVA’s
Act Between                               operating policies regarding its multipurpose tributary projects. However,
Operating Priorities                      the needs of these various users often conflict and/or compete with each
                                          other, and TVA must attempt to balance its operations to best meet the
and Users’ Interests                      needs of these users while adhering to its operating priorities of flood
                                          control, navigation, and hydropower generation.



                                          Page 36                                  GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 2
Multipurpose Tributary Projects—Why Do
They Exist and How Are They Operated




Lake user groups, such as the Land Owners and Users of Douglas (LOUD)
and the Cherokee Lake Users Association (CLUA), which primarily include
the residents and commercial businesses located on or near the tributary
lakes and those that use the lakes for recreational purposes, have sought
higher lake levels throughout the year. These groups argue that higher lake
levels result in enhanced recreational opportunities, which have become a
growing percentage of the economy in the counties encompassing and
surrounding the lakes. The groups also argue that higher lake levels have a
positive affect on property values. Marina owners on the Douglas and
Cherokee Lakes explained that their business is significantly reduced
starting in late summer as a result of TVA’s lake-level drawdown beginning
on August 1. According to these individuals, many marina owners have
closed down their businesses in the past because of TVA’s policies
impacting lake levels. Local government officials tended to reiterate the
concerns of the lake users and stressed the importance of recreation and
tourism in their local economies. While all of these entities and
organizations recognized that changes made as a result of TVA’s 1990
review did improve lake levels, they all believe that TVA’s policy of
beginning the lake-level drawdown on August 1 has a severe impact on the
tourism and recreational trade in these areas.

Other users, such as a barge operator and a boat manufacturer, had very
different views on TVA’s operation of its integrated system. These users
cited minimum channel levels on the Tennessee River as their primary
concern. TVA’s policy of maintaining an 11-foot channel allows the barge
operator to move products up and down the river and gives the boat
manufacturer the ability to test its newly constructed boats. Any
significant changes to TVA’s operating policies that may have a negative
effect on downstream flow of water could have a significant impact on
such businesses. Still other users, such as TVA’s power distributors, who
purchase TVA’s wholesale power and then resell the power to the retail
consumers, are primarily concerned with any increase in the cost of power
that would result from changes in TVA’s policies impacting lake levels. TVA’s
current policies allow for significant hydroelectric power generation
during the peak periods of power demand, when the power is most
valuable. Any shift in these policies could result in a higher cost of power
for TVA, which could, in turn, increase the cost of power to the power
distributors and retail customers.

Emergency management officials near Chattanooga and businesses
located near TVA projects are concerned with potential increased risks of
flooding that may result from changes in TVA’s operating policies. Officials



Page 37                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 2
Multipurpose Tributary Projects—Why Do
They Exist and How Are They Operated




representing state and federal environmental agencies expressed concerns
with the quality of the water that flows down the Tennessee River through
TVA’s projects. Environmental concerns include the concentration of
wastewater in the river flow that results from industries located along the
river and the change in temperature and/or the level of oxygen in the water
that may affect aquatic habitats in the river.




Page 38                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 3

Multipurpose Tributary Project Operations
Were Changed as a Result of TVA’s 1990
Review
                     Over the past 3 decades, TVA has instituted two sets of significant changes
                     in the way the multipurpose tributary projects are operated. The last
                     change—made in 1990—resulted in a 2-month delay in the annual
                     drawdown of the lake levels at the multipurpose tributary projects during
                     the summer. TVA rejected a number of other lake-level alternatives because
                     they would have resulted in higher systemwide costs of supplying electric
                     power. Of the seven alternatives evaluated at that time, TVA adopted the
                     one having the least cost increase—estimated at an annual average of
                     $2 million (in 1990 dollars). TVA used a complex methodology in estimating
                     the impacts on TVA’s systemwide cost of supplying electric power of
                     lake-level alternatives. However, this methodology did not quantify
                     (1) other cost impacts, such as impacts to flood control and navigation
                     operations, or (2) the economic benefits of the alternatives evaluated.
                     Because market conditions and evaluation methodologies have changed
                     since 1990, we requested that TVA analyze two of the alternatives evaluated
                     in 1990 to illustrate what the potential impacts on TVA’s systemwide cost of
                     supplying electric power could be in the future. This 1999 analysis showed
                     that delaying the drawdown of tributary lakes from August 1 to Labor Day
                     could result in potential increased costs ranging from $0 to $88 million
                     annually depending on various assumptions, with an average estimated
                     increase of $47 million (in 1999 dollars). TVA cautioned that both the 1990
                     and 1999 estimates were subject to a great deal of uncertainty due to
                     future hydrological conditions,1 electricity prices, and other variables.


                     In 1971, TVA conducted a study to modify, if possible, some portions of its
1971 Study Led to    operations to improve recreational uses of TVA’s multipurpose tributary
Higher Winter Lake   projects within the framework of the statutory requirements for flood
Levels               control, navigation, and hydropower generation. As a result of this study,
                     TVA concluded that raising the January 1 target levels and the normal
                     minimum levels2 of nine of its multipurpose tributary projects should
                     provide higher lake levels during the winter in most years.3 According to
                     TVA’s analysis, such changes would not significantly affect its ability to
                     provide flood control during the month of January. TVA also predicted that

                     1
                      Hydrological conditions characterize water volumes and flows that affect TVA’s objectives of flood
                     control, navigation, power supply, etc.
                     2
                      The normal minimum level refers to the lowest lake elevation that the project would be operated at
                     under normal conditions. TVA would not schedule operations below this elevation without the Board
                     of Directors’ approval.
                     3
                      As a result of this study, TVA modified lake levels at Chatuge, Cherokee, Douglas, Fontana, Hiwassee,
                     Norris, Nottely, South Holston, and Watauga. No modifications in operations at Boone Lake were
                     proposed. According to TVA, the original design of the Boone project specified operations at
                     prescribed seasonal elevations.



                     Page 39                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                          Chapter 3
                                          Multipurpose Tributary Project Operations
                                          Were Changed as a Result of TVA’s 1990
                                          Review




                                          higher winter levels would improve the chances of the lakes filling to
                                          higher levels in the spring, thus enhancing recreational opportunities. For
                                          example, target levels for January 1 were increased as much as 10 feet at
                                          Cherokee and Hiwassee. In addition, the normal minimum level at Douglas
                                          was raised 20 feet while increases in the other lakes ranged from 30 to 100
                                          feet. Table 3.1 highlights the changes TVA implemented in 1971.


Table 3.1: Changes Made in 1971 to Multipurpose Tributary Lake Levels
In feet above sea level

Multipurpose                 Changes to January 1st target levels                        Changes to normal minimum levels
tributary lakesa       Pre-1971 level         1971 level           Increase         Pre-1971 level           1971 level      Increase
Chatuge                        1,910                1,912                   2                 1,860               1,905           45
Cherokee                       1,020                1,030                  10                    980              1,020           40
Douglas                          935                 940                    5                    920                 940          20
Fontana                        1,615                1,620                   5                 1,525               1,580           55
Hiwassee                       1,455                1,465                  10                 1,415               1,450           35
Norris                           978                 985                    7                    930                 960          30
Nottely                        1,743                1,745                   2                 1,690               1,735           45
South Holston                  1,702                1,702                   0                 1,616               1,675           59
Watauga                        1,934                1,940                   6                 1,815               1,915          100
                                          a
                                          These lakes are the only multipurpose tributary lakes that were changed in 1971.

                                          Source: TVA.




                                          In the late 1980s, TVA undertook an operation and planning review of its
1990 Review Resulted                      system of projects. Among the reasons for this review was to reexamine
in Latest Changes in                      the trade-offs TVA makes in balancing the various purposes served by the
TVA’s Policies                            multipurpose tributary projects. This review, which examined seven
                                          lake-level alternatives, ultimately resulted in about a 2-month delay of the
Impacting Lake Levels                     annual lake-level drawdown from Memorial Day until August 1. Several
                                          factors influenced the changes to TVA’s policies impacting lake levels.
                                          These factors included recommended year-round minimum flows on the
                                          Tennessee River that would improve water quality and support navigation
                                          and the effect of operating changes on TVA’s ability to generate
                                          hydropower.


Reasons Why the 1990                      In September 1987, TVA’s Board of Directors authorized a review to
Review Was Undertaken                     determine whether the operating priorities for its projects that had been




                                          Page 40                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                      Chapter 3
                      Multipurpose Tributary Project Operations
                      Were Changed as a Result of TVA’s 1990
                      Review




                      set out in the TVA Act over 50 years earlier still made sense given the
                      changes that had taken place in the Tennessee Valley since the 1930s.
                      Although the act directed that the system of projects be managed primarily
                      for navigation and flood control and then for power generation, there was
                      a recognition in the late 1980s that other benefits had also become
                      important to residents of the Valley. Specifically, concerns had been raised
                      that TVA’s policies often resulted in prolonged periods without releases
                      from the dams. As a result, the effects on aquatic life could be severe
                      because areas downstream from the projects were essentially dry for
                      periods of time. In addition, public demand for abundant supplies of clean
                      water and for recreation on TVA’s lakes and streams was competing with
                      the demand for the benefits associated with the continued production of
                      low-cost TVA hydropower. By the time the review was authorized, growing
                      numbers of lake users had asked TVA to reexamine the trade-offs it makes
                      in balancing the various purposes when operating its integrated river
                      system.

                      In December 1990, TVA released the results of its work examining lake
1990 Review Results   management policies in a report entitled, Tennessee River and Reservoir
in Delay of Annual    System Operation and Planning Review. In carrying out the work for its
Lake-Level Drawdown   review, TVA followed the National Environmental Policy Act by preparing a
                      detailed statement on the environmental impact of proposed actions that
Date Until August 1   could significantly affect the quality of the human environment. By
                      preparing an environmental impact statement, “TVA sought to ensure that
                      the environmental effects of reservoir operating alternatives were
                      thoroughly investigated and that ample opportunities for pubic review and
                      comment were provided.” Referred to by TVA as its “Lake Improvement
                      Plan,” this review evaluated (1) three alternatives to provide additional
                      minimum flows from TVA dams to improve reservoir releases downstream
                      and (2) seven alternatives to stabilize lake levels by delaying the
                      drawdown of lake levels until August 1 or later.

                      As a result of TVA’s analyses, the 1990 review recommended that (1) TVA
                      increase minimum flow requirements from mainstream and tributary
                      projects and increase dissolved oxygen levels in the releases from 16 of its
                      dams and (2) maintain summer target levels in 10 multipurpose tributary




                      Page 41                          GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                         Chapter 3
                                         Multipurpose Tributary Project Operations
                                         Were Changed as a Result of TVA’s 1990
                                         Review




                                         projects until August 1st(Alternative 1).4 On the basis of TVA’s analyses,
                                         these alternatives were preferred because they provided the most benefits
                                         at the least cost to TVA. Table 3.2 shows the seven lake-level alternatives
                                         TVA evaluated.



Table 3.2: Lake-Level Alternatives Analyzed by TVA During 1990 Review
                                           Drawdown restricted from Memorial Day
Alternative                                to                                                Projects affected
1                                        August 1 for 10 projects                            Blue Ridge, Chatuge, Cherokee, Douglas,
                                                                                             Fontana, Hiwassee, Norris, Nottely, South
                                                                                             Holston, and Watauga
1A                                       August 1 for 7 projects and October 1 for 3         October 1 for the Cherokee, Douglas, and
                                         Knoxville area projects                             Norris projects and August 1 for the other
                                                                                             7 projects
1B                                       August 1 for 8 projects and October 1 for 2         October 1 for the South Holston and
                                         Tri-Cities area projects                            Watauga projects and August 1 for the
                                                                                             other 8 projects
1C                                       August 1 for 9 projects and October 1 for 1         October 1 for the Fontana project and
                                         project                                             August 1 for the other 9 projects
1D                                       August 1 for 6 projects and October 1 for 4         October 1 for the Blue Ridge, Chatuge,
                                         Hiwassee basin projects                             Hiwassee, and Nottely projects and
                                                                                             August 1 for the other 6 projects
2                                        Labor Day for 10 projects                           Blue Ridge, Chatuge, Cherokee, Douglas,
                                                                                             Fontana, Hiwassee, Norris, Nottely, South
                                                                                             Holston, and Watauga
3                                        October 31 for 10 projects                          Blue Ridge, Chatuge, Cherokee, Douglas,
                                                                                             Fontana, Hiwassee, Norris, Nottely, South
                                                                                             Holston, and Watauga
                                         Source: TVA.



                                         In TVA’s evaluation of lake-level alternatives, TVA estimated that changes in
                                         the operation of its system of projects would increase TVA’s systemwide
                                         cost of supplying electric power. On the basis of TVA’s assumption of 60

                                         4
                                          During the review, the 10 TVA projects analyzed for lake-level changes included 9 multipurpose
                                         tributary projects that were subject to significant summer drawdown—Chatuge, Cherokee, Douglas,
                                         Fontana, Hiwassee, Norris, Nottely, South Holston, and Watauga. The Blue Ridge project—which is a
                                         single-purpose power project—was also included because it has an annual drawdown cycle similar to
                                         multipurpose tributary projects. The remaining four multipurpose projects—Boone, Melton Hill,
                                         Tellico, and Tims Ford—not included in the review were excluded for various reasons. Boone was
                                         excluded because its original design included its operation at prescribed seasonal elevations that
                                         result in a constant lake elevation from Memorial Day through Labor Day. Melton Hill does not have an
                                         annual drawdown; it is operated in a fixed range of about 793 feet to 795 feet. Tellico, which is
                                         connected by an ungated canal to Fort Loudoun Lake, has a lake elevation essentially the same as Fort
                                         Loudoun—a multipurpose main river project. Because Fort Loudoun is targeted to reach its summer
                                         lake level by April 15 and its drawdown does not begin until November 1, Tellico has a flat summer
                                         lake level until November 1. Tims Ford, by design and original project allocation, has always been
                                         operated with a minimum summer lake elevation of 883 feet, which extends until October 15.



                                         Page 42                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 3
Multipurpose Tributary Project Operations
Were Changed as a Result of TVA’s 1990
Review




hours of peak generation per week during the summer, a delay in the
drawdown of lake levels would limit TVA’s ability to generate hydropower
during those peak hours. To maximize the value of its hydropower, TVA
attempts to schedule hydropower generation during the periods of
greatest demand to offset the use of other, more expensive, sources of
power generation. According to TVA’s records, the highest and longest
periods of peak demand occur during the summer. For example, the daily
summer peak demands last 8 or more continuous hours a day. Given that
TVA’s power resources are constant, limiting TVA’s ability to generate
hydropower during such peak periods may force TVA to use more
expensive sources of generation to meet the demand. While much of the
water that could have been used to generate peak power could eventually
be shifted to offpeak times later in the year, the demand for power during
these offpeak periods is lower than the demand for power during peak
periods. As a result, the net effect of this shift in hydroelectric power
production would generally result in a net increase in TVA’s systemwide
cost of supplying electric power.

TVA made some adjustments to Alternative 1 that made it the most
attractive alternative examined in the 1990 review. For example, TVA
reduced the effects of lake-level changes on its ability to generate
hydropower during peak periods in the summer by developing “sloping”
target lake levels for the multipurpose tributary projects. The sloping
target levels allowed TVA to fill the lakes above the August 1 summer target
levels. By doing so, TVA could conduct a limited drawdown, referred to as a
“restricted” drawdown,5 of the lake levels during June and July and use
this additional water to generate hydroelectric power during periods of
peak power demand in these 2 months, while still meeting the target lake
levels on August 1, when “unrestricted” drawdown could begin. However,
applying sloping target levels to other alternatives, such as the Labor Day
alternative, prevented the use of enough water for hydropower generation
during peak demand periods in June, July, and August because water
would have to be reserved to provide minimum flows throughout the
extended summer period. In addition, TVA recommended that it provide
minimum flows using turbine pulsing, which allows TVA to produce power




5
 TVA officials stated that this “restricted” drawdown in June and July can total several feet of
lake-level elevation. Examples cited were 11 feet at Cherokee, 10 feet at Watauga and Fontana, and 8
feet at South Holston.



Page 43                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 3
Multipurpose Tributary Project Operations
Were Changed as a Result of TVA’s 1990
Review




while releasing the water.6 Other options of providing minimum flows
from the projects would result in the total loss of power generation from
the released water.

Table 3.3 shows the effects of the changes on the August 1 lake levels of
the 10 multipurpose tributary projects considered in the 1990 review. To
illustrate what the current lake levels are, we show in appendix II the
monthly minimum and maximum lake levels that occurred during calendar
year 1998 at the multipurpose tributary projects, and in table 3.4 we
highlight the total elevation change that occurred during calendar year
1998 at these projects. In addition, table 3.5 shows the key target lake
levels during the year.




6
 Turbine pulsing refers to a process whereby water is released through the turbines of a TVA tributary
project by operating the turbines for short periods of time throughout the day. For example, TVA
recommended that at six projects it would “pulse” the turbines for 30 to 60 minutes every 4 hours
when the turbines otherwise would not be operating. While operating the turbines in this manner, TVA
also has the ability to generate hydropower.



Page 44                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                       Chapter 3
                                       Multipurpose Tributary Project Operations
                                       Were Changed as a Result of TVA’s 1990
                                       Review




Table 3.3: Effect of 1990 Changes to
Multipurpose Tributary Median Lake     In feet above sea level
Levels                                                                                     Median lake levels (Aug. 1)a
                                       Multipurpose tributary                   Prior to 1990
                                       project                                        reviewb After 1990 reviewc                     Change
                                       Blue Ridge                                      1,680.2                1,683.1                       +2.9
                                                                                               d                       d
                                       Boone                                                                                                N/A
                                       Chatuge                                         1,920.5                  1,924                       +3.5
                                       Cherokee                                        1,054.1                1,061.5                       +7.4
                                       Douglas                                           984.4                  990.9                       +6.5
                                       Fontana                                         1,677.9                1,696.6                   +18.7
                                       Hiwassee                                        1,514.1                1,517.2                       +3.1
                                                                                               d                       d
                                       Melton Hill                                                                                          N/A
                                       Norris                                            1,005                1,013.2                       +8.2
                                       Nottely                                         1,767.8                1,771.2                       +3.4
                                       South Holston                                   1,718.2                  1,722                       +3.8
                                                                                               d                       d
                                       Tellico                                                                                              N/A
                                                                                               d                       d
                                       Tims Ford                                                                                            N/A
                                       Watauga                                         1,941.9                1,950.7                       +8.8
                                       a
                                        The term “median” refers to the middle number in an ordered series of numbers. If the series
                                       contains an even number of items, then the median refers to the number midway between the two
                                       middle numbers in the ordered series of numbers.
                                       b
                                           Based on data from 1972 through 1990.
                                       c
                                       Based on data from 1991 through 1998.
                                       d
                                           The Boone, Melton Hill, Tellico, and Tims Ford projects were not included in TVA’s 1990 study.

                                       Source: TVA.




                                       Page 45                                  GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                           Chapter 3
                                           Multipurpose Tributary Project Operations
                                           Were Changed as a Result of TVA’s 1990
                                           Review




Table 3.4: Multipurpose Tributary Lake Levels—Minimum and Maximum Level in Calendar Year 1998
In feet above sea level

Multipurpose         Minimum elevation in calendar year 1998      Maximum elevation in calendar year 1998                 Total elevation
tributary lake                 Elevation                 Month                Elevation                      Month                change
Blue Ridgea                     1,621.38              November                  1,689.07                        April                67.69
Boone                           1,351.35                January                 1,383.80                        May                  32.45
Chatuge                         1,911.08             December                   1,926.31                       June                  15.23
Cherokee                        1,026.37             December                   1,071.76                       June                  45.39
Douglas                           940.29             December                     995.36                        April                55.07
Fontana                         1,625.68             December                   1,705.34                       June                  79.66
Hiwassee                        1,460.31             December                   1,524.30                       June                  63.99
Melton Hill                       790.10                  March                   796.44                        April                 6.34
Norris                            977.02                January                 1,030.38                        April                53.36
Nottely                         1,740.84             December                   1,778.09                       June                  37.25
South Holston                   1,689.40                January                 1,732.29                        April                42.89
Tellico                           807.45               February                   815.09                        April                 7.64
           a
Tims Ford                         856.01                January                   890.14                       June                  34.13
Watauga                         1,930.63                January                 1,960.69                        April                30.06
                                           a
                                         TVA said that the elevations for Blue Ridge beginning in September were abnormally low due to
                                        a special drawdown for maintenance inspection that is conducted once every 5 years. In
                                        addition, TVA said that the elevations for Tims Ford in January and February were abnormally low
                                        due to a special drawdown for repairs.

                                        Source: TVA.




                                        Page 46                              GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                         Chapter 3
                                         Multipurpose Tributary Project Operations
                                         Were Changed as a Result of TVA’s 1990
                                         Review




Table 3.5: Multipurpose Tributary Lake
Level Targets Throughout the Year        In feet above sea level
                                                                                                             Minimum targeted summer
                                         Multipurpose                   Flood control level                           levels
                                         tributary lake               January 1             March 15                June 1             August 1
                                         Blue Ridge                         1,668                1,678                1,685                1,682
                                         Boone                              1,358                1,375                1,382                1,382
                                         Chatuge                            1,912                1,916                1,924                1,923
                                         Cherokee                           1,030                1,042                1,061                1,060
                                         Douglas                              940                  958                  992                   990
                                         Fontana                            1,644                1,644                1,696                1,693
                                         Hiwassee                           1,465                1,482                1,517                1,515
                                                                                   a                    a                    a                    a
                                         Melton Hill
                                         Norris                               985                1,000                1,012                1,010
                                         Nottely                            1,745                1,755                1,772                1,770
                                         South Holston                      1,702                1,713                1,723                1,721
                                         Tellico                              807                  807                  813                   813
                                         Tims Ford                            873                  879                  883                   883
                                         Watauga                            1,940                1,951                1,950                1,949
                                         a
                                          These dates have no set level due to the small fluctuation in lake levels; the full range is from 790
                                         to 796 feet above sea level, with a normal operating range of 793 to 795 feet above sea level.

                                         Source: TVA.




Several Factors Influenced               According to TVA’s analysis, none of the lake-level alternatives considered
the 1990 Changes                         in the 1990 review would have negatively affected flood control and
                                         navigation on the Tennessee River. All of the alternatives required TVA to
                                         reach flood control target levels by January 1, in preparation for the winter
                                         months, where flood risk is the greatest in the Tennessee Valley.
                                         Navigation on the Tennessee River would be supported by recommended
                                         year-round minimum flows, and navigation during the typically dry fall
                                         months would benefit from additional flows from the delayed lake-level
                                         drawdown. Each of the alternatives that maintained summer target lake
                                         levels at the multipurpose tributary projects would increase recreational
                                         opportunities on these lakes. However, some of the alternatives, such as
                                         the alternative that delayed the annual lake-level drawdown at the
                                         multipurpose tributary projects until October 31 (Alternative 3), could
                                         have increased the risk of flooding of the Ohio and Mississippi Rivers,
                                         which connect with the Tennessee River near Paducah, Kentucky.
                                         According to TVA, no rigorous flood risk analysis was performed by TVA as




                                         Page 47                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
    Chapter 3
    Multipurpose Tributary Project Operations
    Were Changed as a Result of TVA’s 1990
    Review




    part of the evaluation of Alternatives 2 and 3 because these alternatives
    were the most expensive and therefore studied the least. The increased
    flooding risk results from TVA’s attempts to draw down its lakes to their
    January 1 levels in a 2-month period as opposed to starting the lake-level
    drawdown earlier in the summer. TVA’s analysis also concluded that the
    October 31 alternative could have negative effects on navigation on the
    Ohio and Mississippi Rivers during dry years.

    Also, according to TVA, Alternative 1 was selected to maintain summer
    target lake levels until August 1 because it provided the most benefits at
    the least cost and because this alternative did not have a significant effect
    on flood control or navigation efforts on either the Tennessee River or the
    Ohio and Mississippi Rivers. The reservoir release alternative implemented
    by TVA increased minimum flows from its mainstream and tributary
    projects and increased dissolved oxygen levels in the releases from 16 of
    its projects. TVA said that the following benefits would be realized as a
    result of the changes implemented from the 1990 review:

•   Minimum flows and aeration of releases would recover over 180 miles of
    aquatic habitat and improve levels of dissolved oxygen in over 300 miles of
    river.
•   Higher summer lake levels would increase lake recreation visitation by
    approximately 21 percent.
•   Scenic views would be improved.
•   Opportunities for tourism and second home development would be
    increased.
•   Reservoir fisheries, due to increased survival of young fish, would be
    improved.
•   Water depth for commercial navigation on the lower Ohio and Mississippi
    Rivers would be increased during September and October, the months of
    lowest flow.

    Another significant factor that TVA considered in its selection of
    alternatives impacting lake levels and reservoir releases was the effect of
    any operating changes on TVA’s ability to generate hydropower. TVA
    attempts to schedule releases of water through its hydropower facilities at
    the multipurpose tributary projects during periods of greatest power
    demand to offset the use of other, more expensive, sources of power. The
    greatest and longest peak power demand periods occur during the
    summer. Any shift in the availability of water for hydropower generation
    due to policies impacting lake levels could restrict TVA’s ability to generate
    hydropower during these peak periods. If a portion of TVA’s hydropower



    Page 48                          GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                        Chapter 3
                        Multipurpose Tributary Project Operations
                        Were Changed as a Result of TVA’s 1990
                        Review




                        generation is delayed until later in the year, TVA must meet the peak power
                        needs by using more expensive forms of generation, which would increase
                        TVA’s cost to produce power. On the basis of analyses using a complex set
                        of models and methodologies, TVA estimated that the lake-level alternative
                        that maintained summer target lake levels until August 1 had the smallest
                        cost increase impact to the power program.


                        The estimated impact to TVA’s systemwide cost of supplying electric power
TVA’s 1990 Review       resulting from delaying the lake-level drawdown of the multipurpose
and Its 1999 Analysis   tributary projects in the summer was a major factor influencing TVA’s
of Two Lake-Level       selection among the seven lake-level alternatives considered. TVA
                        estimated that delaying the lake-level drawdown at its multipurpose
Alternatives Both       tributary projects from Memorial Day to August 1 or later would increase
Estimate Increases in   its systemwide annual average cost of supplying electric power by
                        $2 million to $93 million (in 1990 dollars), depending on the alternative.
TVA’s Systemwide        TVA’s 1990 cost estimation involved a complex methodology using
Cost of Supplying       hydrologic and electric supply and demand computer models and
Electric Power          extensive data. Our evaluation of TVA’s 1990 efforts included a review of
                        the available documentation supporting the key elements of its
                        calculations of the impacts on its systemwide cost of supplying electric
                        power. Because market conditions and evaluation methodologies have
                        changed since 1990, we requested that TVA update its estimated
                        systemwide cost impacts of implementing lake-level alternatives similar to
                        two of the alternatives considered in 1990—Alternative 1A (an additional
                        2-month drawdown delay for three of the projects) and Alternative 2 (an
                        additional 1-month drawdown delay for all of the projects).

                        TVA’s March 1999 analysis showed that the two alternatives would increase
                        TVA’s systemwide cost of supplying electric power by an estimated annual
                        average cost of $14 million and $47 million, respectively (in 1999 dollars).
                        TVA cautioned that both the 1990 and 1999 estimates were subject to a
                        great deal of uncertainty. Some elements of uncertainty were built into the
                        1999 estimates. For example, the estimated $14 million increase could be
                        as low as $(-) 2 million (negative $2 million—that is, benefits, not costs) or
                        as high as $33 million, depending on hydrological conditions, future
                        electricity prices, and other factors. However, TVA said that there were
                        other important elements of uncertainty that it did not attempt to, and
                        could not easily, model. TVA has emphasized that a comprehensive
                        evaluation of proposed policy changes would go well beyond estimating
                        the effect of the changes on TVA’s systemwide cost of supplying electric
                        power and would also have to consider important equity implications of



                        Page 49                          GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
              Chapter 3
              Multipurpose Tributary Project Operations
              Were Changed as a Result of TVA’s 1990
              Review




              proposed changes. While a complete evaluation of the cost-estimation
              methodologies TVA used was beyond the scope of our work, the general
              approach TVA used in 1990 and 1999 appeared to be reasonable.

              The analyses of lake-level alternatives conducted by TVA in 1990 and 1999
              concentrated on developing estimated costs of various alternatives. TVA’s
              1990 study recognized and discussed various benefits and costs—in
              addition to the systemwide costs of supplying electric power—of changes
              to lake levels, such as the impacts to flood control and navigation efforts.
              However, TVA’s estimates related only to the systemwide costs of supplying
              electric power. TVA officials explained that TVA does not currently possess
              adequate tools and methodologies to develop such quantitative
              assessments. Similarly, TVA’s analyses did not attempt to quantify, in
              monetary terms, the potential benefits that may occur as a result of the
              lake-level alternatives. For example, TVA explained that it does not have
              adequate mechanisms to capture visitation statistics at its lakes or to
              estimate the monetary effects of recreation and tourism on local
              economies. In its 1990 review, TVA based its analyses of recreation visits at
              TVA lakes on an inventory of access facilities, staff judgment, and
              interviews with facility operators. While TVA concluded that lake-level
              Alternative 1 would increase recreation visits by 21 percent, TVA did not
              attempt to quantify this estimated visitor increase into monetary terms.7

              We provide additional information in appendix III about the
              methodologies used and the results of TVA’s 1990 review and its 1999
              analysis.


              The key changes resulting from TVA’s 1990 review were to improve water
Conclusions   quality and aquatic habitat and to extend the recreation season on TVA
              lakes by delaying the annual drawdown of the multipurpose tributary
              project lake levels for an additional 2 months during the summer. The
              impact on TVA’s systemwide cost of supplying electric power resulting
              from the delay in the annual lake-level drawdown was the major factor
              influencing these changes. TVA’s analyses show that even when only
              looking at these cost impacts, there is still a level of uncertainty associated
              with the results because of the assumptions being made. In addition, TVA
              did not attempt to quantify potential benefits that may result from

              7
               TVA officials explained that a panel of nine external reviewers for the 1990 study advised TVA to fully
              describe the benefits and costs of each alternative, including estimated power costs, and focus public
              participation during the NEPA review of the draft environmental impact statement on which
              alternative was best rather than on how monetary estimates were calculated for benefits for which no
              market value exists.



              Page 50                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 3
Multipurpose Tributary Project Operations
Were Changed as a Result of TVA’s 1990
Review




increased recreation or tourism by maintaining summer lake levels longer.
Furthermore, neither TVA’s 1990 review nor the 1999 analysis TVA
performed at our request attempted to quantify the potential cost effects
on other aspects of TVA’s operations, such as flood control and navigation.
Therefore, decisions about how to regulate the levels of the lakes have not
been based on a complete evaluation of the overall costs and benefits of
such actions.




Page 51                          GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 4

Little Has Changed in Multipurpose
Tributary Project Operations Since the 1990
Review Despite Requests From Users
                         Although TVA continued to receive requests from individuals and
                         organizations for changes to the multipurpose tributary projects after the
                         1990 review, little has changed in how TVA operates these projects. TVA
                         raised concerns about the growing number of requests from lake users
                         asking TVA to analyze multiple lake-level alternatives for individual lakes.
                         TVA decided that a “piecemeal” approach to analyzing the issues involved
                         raised questions of fairness in how each of the multipurpose tributary
                         projects would be treated within the TVA system. As a result of the
                         numerous requests, in March 1997, TVA decided to establish a 4-year
                         moratorium on changes to policies impacting lake levels. TVA cited several
                         reasons for taking this action, including allowing TVA time to evaluate how
                         studies on policies impacting lake levels should be evaluated in the future.
                         Since the moratorium’s implementation, one group of lake users has
                         submitted a request to TVA for changes to policies impacting lake levels. In
                         addition, TVA has commented on two studies discussing the potential
                         economic benefits resulting from maintaining higher lake levels later in
                         the year at Cherokee and Douglas Lakes in Tennessee, and Blue Ridge,
                         Chatuge, and Nottely Lakes located in northern Georgia. TVA’s comments
                         on these two studies were critical of several aspects of the studies,
                         including the scope, methodology, assumptions, and data used to estimate
                         the economic benefits. TVA also said that the two studies lacked a proper
                         recognition of the multipurpose roles served by TVA’s system of projects
                         and how changes in lake levels may impact its operations and
                         management of the entire Tennessee River system. Our examination of the
                         two studies identified limitations in the studies’ methodologies and scope.
                         For example, both studies are limited in their scope because the benefits
                         they estimated are only those associated with increased lake visitation
                         pertaining to a few counties adjacent to the lakes.


                         Despite the changes made to its policies impacting lake levels earlier this
In 1997 TVA Decided      decade, TVA has continued to receive a number of requests to make further
to Establish 4-Year      changes. TVA ultimately decided in March 1997 to implement a 4-year
Moratorium on            moratorium on making any further changes to these policies. There were a
                         number of factors influencing TVA’s decision, including a belief that the
Changes to Project       moratorium would minimize the public’s perception of favoritism for any
Operations               particular lake and would also allow time for TVA to evaluate how studies
                         on policies impacting lake levels should be evaluated in the future.


Requests for Changes     After the 1991 Lake Improvement Plan was implemented, requests for
Made From 1990 Through   changes to TVA’s lake-level policies slowed for a year or two but began
1997                     again in 1993. According to TVA, constituents were no longer satisfied with


                         Page 52                     GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 4
Little Has Changed in Multipurpose
Tributary Project Operations Since the 1990
Review Despite Requests From Users




the changes made in 1991, or new constituents were not aware of the
changes that had been made. In these instances, when TVA received
requests for changes, TVA’s staff would acknowledge these requests and
meet with interested individuals and/or organizations. Requests from
individuals, although acknowledged, were seldom evaluated in detail.
Requests from organizations which, after a limited review, would tend to
have a low or negligible cost with few negative impacts, were further
evaluated by an internal inter-disciplinary team. A few of these were
implemented, either on a permanent or trial basis. TVA staff would decide
which requests deserved further attention and then perform some level of
study to be able to answer questions about potential impacts of the
changes. In at least two instances, TVA made offers to “sell” higher lake
levels to two groups (the state of North Carolina for Fontana Lake and a
group of concerned citizens at Blue Ridge Lake). Other studies were
underway for Boone Lake and the North Georgia lakes when the
moratorium was implemented.

By March of 1997, several requests for changes to policies impacting lake
levels had been submitted to TVA. For example, (1) TVA had completed a
preliminary study that examined the power and flood control impacts of
extending Boone Lake’s level later into the fall; (2) TVA had met with the
Mountain Lakes Study Committee, which was performing an economic
analysis of 17 alternative drawdown scenarios for the Blue Ridge, Chatuge,
and Nottely Lakes; (3) the Cherokee Lake Users Association had met with
TVA and proposed changes to the annual drawdown policy at Cherokee
Lake; (4) individual users at South Holston and Watauga Lakes were
requesting changes in policy at those lakes; and (5) there were
miscellaneous requests for changes in operation at other TVA tributary and
main river projects.

TVA staff had performed analyses for Boone Lake, which indicated that the
impacts on TVA’s systemwide cost of supplying electric power associated
with the requested changes were relatively small, with a net present value
of less than $1 million. TVA estimated that increased systemwide cost of
supplying electric power associated with the requested changes at Boone
Lake was much less than for other TVA lakes analyzed in the past, primarily
because the changes in the lake levels during the year at Boone Lake were
smaller in comparison to other lakes, and TVA already extended the
summer target lake level at Boone Lake until Labor Day.1 As a result, TVA
would not need to shift power production at Boone Lake from the peak

1
 The original design of the Boone project included its operation at prescribed seasonal elevations. For
example, the lake is operated at a constant elevation of about 1,382 feet above sea level from Memorial
Day through Labor Day.



Page 53                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                            Chapter 4
                            Little Has Changed in Multipurpose
                            Tributary Project Operations Since the 1990
                            Review Despite Requests From Users




                            summer months to the fall. In addition, the flood impact analyses indicated
                            that based on historic data, flood control at Boone Lake would not be
                            impacted. However, TVA indicated that potential storms would have an
                            impact on the frequency of floods downstream from Boone Lake.


Factors Influencing TVA’s   TVA became concerned that more and more users were requesting studies
March 1997 Moratorium       for the lakes they used, resulting in an analysis of the system on a
                            piecemeal basis. To TVA, this raised a “fairness” issue of treating these
                            lakes differently in the TVA system. Of particular concern to TVA was the
                            relatively low impact that the requested changes at Boone Lake would
                            have on TVA’s systemwide cost of supplying electric power. TVA believed
                            that the implementation of these changes would give even more favoritism
                            to a lake that already had high lake levels envied by users at other
                            tributary lakes, while also promoting a “first come/first served” attitude
                            to the lake users. For example, if the earliest request for lake-level changes
                            were granted for an individual lake, it may make changes to other lakes
                            more expensive when implemented later. Additional concerns cited by TVA
                            included: (1) difficulties with the current methodology to systematically
                            and objectively address possible flood concerns; (2) the great uncertainty
                            surrounding the future of the electric utility industry due to deregulation
                            and restructuring; (3) implementing changes at any lake would raise the
                            expectation for changes at all lakes; and (4) although the cost of Boone
                            Lake changes were relatively small, the lake users were not prepared to
                            reimburse TVA for the increased systemwide cost of supplying electric
                            power, and TVA was not prepared to pass this cost to the ratepayer.

                            In March 1997, TVA’s Board of Directors agreed with the Executive
                            Committee’s determination that a moratorium on changes to policies
                            impacting lake levels was the proper course of action.2 According to TVA,
                            the 4-year moratorium would position it better for future competition in
                            the electric utility industry, by retaining the current operating flexibility
                            afforded by its hydroelectric power facilities. TVA also believed that the
                            moratorium would minimize the public’s perception of favoritism for any
                            particular lake on the system and would allow time for TVA staff to
                            evaluate how studies on policies impacting lake levels should be evaluated
                            in the future.




                            2
                             The Board of Directors appointed TVA’s nine senior officers to the Executive Committee. This
                            committee meets regularly to coordinate TVA’s activities across organizational lines and to ensure that
                            matters requiring Board approval are appropriately referred.



                            Page 54                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                     Chapter 4
                                     Little Has Changed in Multipurpose
                                     Tributary Project Operations Since the 1990
                                     Review Despite Requests From Users




                                     On December 9, 1997, the President of LOUD, requested that TVA review and
December 1997                        consider the organization’s alternative policy to the annual drawdown at
Request for Changes                  Douglas Lake. The LOUD proposal, compared with TVA’s existing policy,
to Policies Impacting                would permit a slower summer drawdown period and a higher January 1
                                     target lake level. LOUD believes that TVA’s existing policy requiring a
Lake Levels From                     elevation of 940 feet above sea level at January 1 (the lowest target level
Users of Douglas Lake                during the year) is too restrictive. According to LOUD, a proposed level of
                                     956 feet above sea level at March 1 (the proposed lowest target level
                                     during the year) is more realistic. LOUD also believes that its proposal
                                     would (1) provide higher lake levels for the late summer, fall, and winter
                                     seasons; (2) present a more scenic view of the lake in the winter; and
                                     (3) increase the probability of filling the lake in the spring. Table 4.1
                                     captures the differences in lake levels on the first day of each month
                                     between LOUD’s proposal and TVA’s median lake levels for the past 8 years.

Table 4.1: Comparison of Douglas
Lake Monthly Elevation               In feet above sea level
Differences—LOUD’s Alternative vs.                                                             TVA’s median
TVA’s Median Lake Levels             1st day of month               LOUD alternative             lake levelsa      Difference
                                     January                                        970                      944         + 26
                                     February                                       963                      947         + 16
                                     March                                          956                      957           -1
                                     April                                          972                      974           -2
                                     May                                            985                      990           -5
                                     June                                           994                      994            0
                                     July                                           993                      994           -1
                                     August                                         991                      991            0
                                     September                                      989                      985          +4
                                     October                                        985                      972         + 13
                                     November                                       981                      959         + 22
                                     December                                       976                      951         + 25
                                     a
                                     Based on data from 1991 through 1998 for the first day of each month.

                                     Source: LOUD and TVA.




                                     Page 55                             GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                       Chapter 4
                       Little Has Changed in Multipurpose
                       Tributary Project Operations Since the 1990
                       Review Despite Requests From Users




                       Prior to LOUD’s formal December 9, 1997, request to TVA, another
Study of the           group—the Cherokee Lake Users Association—had been making requests
Economic and Fiscal    for TVA to increase the lake levels at Cherokee Lake. After the LOUD
Consequences of        proposal had been formally made to TVA, the LOUD and CLUA groups sought
                       to develop additional information that would show the potential economic
TVA’s Draw-Down of     benefits resulting from higher lake levels at both the Cherokee and
the Cherokee and       Douglas Lakes. Towards this end, both organizations were instrumental in
                       ensuring that the state of Tennessee and local governments within a
Douglas Lakes          six-county region surrounding both lakes would fund a study exploring
                       “the benefits which might accrue to residents of and visitors to a
                       six-county region surrounding Douglas and Cherokee Lakes, should TVA
                       alter its lake level policy.”3 The study, Economic and Fiscal Consequences
                       of TVA’s Draw-Down of Cherokee and Douglas Lakes (henceforth, the
                       Cherokee and Douglas Study), was completed in October 1998 and was
                       conducted by the University of Tennessee’s Center for Business and
                       Economic Research. TVA received copies of the report and met with the
                       authors of the report, as well as LOUD and CLUA representatives in
                       January 1999. The meeting was scheduled by TVA at the request of LOUD for
                       the purpose of sharing comments and answering questions about the
                       report.

                       The study used several methodologies to estimate the economic impacts
                       of changes to lake levels on areas bordering the two lakes. The study
                       concludes that positive economic and fiscal impacts would result from
                       proposed changes to policies impacting lake levels. LOUD representatives
                       and TVA identified weaknesses in the study. In addition, in our review of
                       the Cherokee and Douglas Study, we have noted some limitations in the
                       study’s scope. We provide additional information in appendix IV about this
                       study, including both TVA’s and our comments on the study.


                       About the same time that LOUD was making its formal proposal requesting
Mountain Lakes Study   TVA to change its operating policy at Douglas Lake, another group, the
Committee Report on    Mountain Lakes Study Committee published a report entitled, The
Georgia Lakes          Economic Impact of Alternate TVA Lake Management Policies (henceforth,
                       the Georgia Mountain Lakes Study). This report focuses on the areas
                       surrounding the three TVA lakes in northern Georgia—Blue Ridge, Chatuge,
                       and Nottely. The Georgia Mountain Lakes Study estimated very favorable
                       benefit-cost ratios for delaying the drawdown of lake levels at Blue Ridge,
                       Nottely, and Chatuge Lakes in northern Georgia from August 1 to
                       October 1. In addition to benefit-cost estimates, the study focuses on

                       3
                        The six counties are: Cocke, Grainger, Hamblen, Hawkins, Jefferson, and Sevier.



                       Page 56                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
              Chapter 4
              Little Has Changed in Multipurpose
              Tributary Project Operations Since the 1990
              Review Despite Requests From Users




              options for funding TVA’s increased cost to its power program due to
              changes to policies impacting lake levels.

              TVA criticized the Georgia Mountain Lakes Study on methodological
              grounds. In our view, the study’s estimates of costs and benefits of
              lake-level alternatives does not conform with recommended federal
              guidelines for the evaluation of major actions on federal water projects.
              We provide additional information in appendix V about this study, as well
              as TVA’s comments and our comments on the study.


              TVA implemented the March 1997 moratorium on making any changes to
Conclusions   policies impacting individual lake levels because of concerns about
              fairness in how each project would be treated within TVA’s integrated
              system. The moratorium, however, has not stopped groups of lake users
              from submitting studies to TVA showing what they perceive as the
              economic benefits that would accrue to the local areas surrounding the
              lakes. Although the results of the studies can be questioned in a number of
              areas, the studies do show that lake users are becoming more concerned
              about identifying and documenting benefits from maintaining higher lake
              levels at certain TVA lakes. We found that the (1) benefit estimations used
              in the studies are not directly comparable with TVA’s estimates of its
              systemwide costs of supplying electric power due to differences in the
              scope of the analyses and (2) studies did not include any detailed
              evaluation of the costs of changes to policies impacting lake levels; the
              studies concentrated on economic benefits to the regions surrounding the
              affected lakes. According to federal guidelines, a comprehensive
              examination of both costs and benefits is recommended when conducting
              analyses of actions related to federal water projects.




              Page 57                          GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 5

Formal and Continuing Communication
Process With the Public and Other
Stakeholders Needed in Future Lake-Level
Study
                      While keeping the 4-year moratorium in place, TVA has recognized that
                      further study of the multipurpose tributary projects is warranted. To
                      pursue this issue, TVA created an internal lake-level policy task force. In
                      July 1998, the task force reported that a reevaluation of policies impacting
                      lake levels should be initiated within the next 2 to 4 years. TVA has
                      budgeted funds to address specific needs that would be required to
                      support future reevaluation efforts. TVA also needs to decide what
                      evaluation option should be pursued under the National Environmental
                      Policy Act guidelines when assessing the environmental effects of any
                      proposed changes to TVA’s policies impacting lake levels.


                      Despite the implementation of its March 1997 moratorium, TVA has
TVA Recognizes That   continued to face ever increasing public concerns about its lake-level
Various Lake-Level    management policies. According to TVA, concerns have been received from
Issues Need to Be     organized user groups and individual constituents and have been
                      expressed in many ways, including calls and letters, referrals through
Addressed             political staffs, and also through public forums such as newspaper
                      editorials. In reacting to these concerns, TVA’s Resource Group made a
                      February 1998 briefing to the Executive Committee.1 The Resource Group
                      cited four reasons for the briefing: (1) it had been a year since the
                      moratorium was put in place, and a briefing could describe the types of
                      requests and concerns still being directed toward TVA; (2) the membership
                      on the Executive Committee is dynamic, and it was important that all TVA
                      offices and organizations be aware of these continuing concerns; (3) it was
                      important that staff reiterate to the Board what the 1991 Lake
                      Improvement Plan did and did not address in terms of operating policy;
                      and (4) to request that the Executive Committee consider whether it was
                      in TVA’s and its customers’ and constituents’ best interest to consider
                      launching a new comprehensive evaluation of its policies impacting lake
                      levels at that time.

                      Resulting from this briefing, the Lake Level Policy Task Force was created
                      and charged with the responsibility to determine the advisability and
                      necessity of conducting a reevaluation of TVA’s policies impacting lake
                      levels. Representatives from many TVA internal organizations participated
                      on the task force to ensure that all TVA constituents were represented. The
                      task force included no organizations external to TVA. The task force
                      investigated the pros and cons of implementing a new study at that time,

                      1
                       On February 8, 1999, the Resource Group was merged with two other TVA offices—Hydro Operations
                      and Hydro Engineering Services—to form a new organization called River System Operations and
                      Environment. According to TVA, this organizational change was made to strengthen the integration of
                      river system management with all operations.



                      Page 58                              GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                        Chapter 5
                        Formal and Continuing Communication
                        Process With the Public and Other
                        Stakeholders Needed in Future Lake-Level
                        Study




                        as well as estimating the time and expense of such a study. The task
                        force’s results, including recommendations, were summarized in an
                        internal July 1998 report, which was the basis for additional briefings for
                        the Executive Committee in September 1998 and January 1999.


The Task Force          The Lake Level Policy Task Force report recommended that while TVA
Recommended That TVA    should continue its moratorium on any changes impacting lake levels, a
Begin Preparation for   reevaluation of such changes should be initiated within the next 2 to 4
                        years. According to TVA officials, this was a recognition that demands for
Future Reevaluation     changes would continue to increase from lake user constituents, and
                        eventually these changes would likely have to be addressed in a
                        comprehensive study. The task force estimated that the cost to conduct a
                        comprehensive review of TVA’s current policies impacting lake levels
                        would total approximately $8 million and would require 3 to 5 years to
                        complete and five full-time employees with additional support staff from
                        various organizations. The report also concluded that, in order to conduct
                        such a review, TVA needed to (1) refine/develop and apply analytical tools
                        aimed at reevaluating flood risk, the impact of policy changes to its
                        systemwide cost of supplying electric power, and economic benefits
                        related to lake-level changes, and (2) develop and implement a proactive
                        communication plan to increase public understanding of TVA’s integrated
                        river system operations.

                        Through its task force’s efforts, TVA has recognized that it needs to
                        improve its evaluation techniques, given the tools and advanced
                        methodologies currently available. According to TVA, preparatory work in
                        this area includes several items, such as (1) refining power evaluation
                        techniques and tracking the evolving market aspects under deregulation
                        guidelines; (2) refining water resource planning and operation models,
                        such as the Weekly Scheduling Model and RiverWare2 to better simulate
                        operation of the integrated system of projects under postulated alternative
                        operation scenarios; and (3) developing objective economic growth and
                        development analysis methods that better represent impacts that could be
                        expected under alternative lake-level scenarios.

                        Because of the lead time necessary to define these analysis techniques,
                        develop the software, and acquire meaningful data, TVA has determined

                        2
                         RiverWare, according to TVA, is a river basin modeling tool with features for simulating or optimizing
                        operation of the system for multiple purposes over varying time horizons with time steps ranging from
                        hourly to monthly. RiverWare is currently used for daily operations scheduling, and TVA anticipates
                        that the planning aspect of this model will eventually supplement or possibly replace the Weekly
                        Scheduling Model.



                        Page 59                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                       Chapter 5
                                       Formal and Continuing Communication
                                       Process With the Public and Other
                                       Stakeholders Needed in Future Lake-Level
                                       Study




                                       that some consistent effort was warranted to begin making these
                                       improvements so as not to unduly delay analyses in the future if and when
                                       they are undertaken. The task force estimated that such an effort would
                                       cost approximately $2 million and would require between 2 to 3 years to
                                       complete. TVA has budgeted additional funds to support several actions in
                                       preparation for an eventual reevaluation of its policies impacting lake
                                       levels. TVA has allocated $1.5 million for fiscal years 1999 through 2001 for
                                       various activities, including the preliminary evaluation of proposed
                                       changes impacting lake levels and the development of flood risk tools. As
                                       shown in table 5.1, TVA has divided this additional funding under two
                                       ongoing programs: Alternate Operations Evaluation and Flood Risk
                                       Reduction. The funds allocated for fiscal year 1999 have been approved;
                                       however, TVA has not approved the planned budgets for succeeding years.

Table 5.1: Funding Budgeted for TVA
Activities Related to the Lake Level                                         Fiscal year
Policy Task Force’s Recommendations    Funding                      1999               2000          2001
                                       category               (approved)          (budgeted)    (budgeted)            Total
                                       Alternate                $100,000           $200,000       $200,000        $ 500,000
                                       operations
                                       evaluations
                                       Flood risk                200,000             400,000       400,000        1,000,000
                                       reduction
                                       Total                    $300,000           $600,000       $600,000       $1,500,000
                                       Source: TVA.



                                       The additional funding for the Alternate Operations Evaluation Program is
                                       designated for improvements in economic evaluation procedures and/or
                                       data collection support. TVA envisions a 3-year effort to better define
                                       economic analysis techniques and to acquire data that would support
                                       these analyses. As part of this effort, TVA is planning to examine
                                       recreational use at several TVA lakes to begin establishing a database for
                                       future lake visitation evaluations. These improvements were estimated to
                                       cost $500,000 and require about 3 years. Under the Flood Risk Reduction
                                       Program, TVA has budgeted an additional $1 million over the next 3 years
                                       to improve the methodology used to establish flood frequencies and to
                                       objectively determine the impact that changes in operations would have.

                                       In addition to the $1.5 million, TVA provides annual funding to both of these
                                       programs for various activities related to improving their tools and
                                       methodologies for evaluating policies impacting lake levels. Under the
                                       Alternate Operations Evaluation Program, TVA spent $364,000 during fiscal



                                       Page 60                         GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 5
Formal and Continuing Communication
Process With the Public and Other
Stakeholders Needed in Future Lake-Level
Study




year 1998 for continuing annual program activities, such as (1) preliminary
evaluations of proposed changes impacting lake levels; (2) staff support
for TVA’s land and water management divisions to promote working
together to increase internal knowledge on how land and water issues are
related; (3) support for staff to maintain expertise in system evaluation
tools, such as the Weekly Scheduling Model, and to improve the reliability
and accuracy of such tools through minor modifications; and (4) the
development of new evaluation tools to complement or eventually replace
existing tools for planning river operations. TVA has allocated $380,000 for
fiscal year 1999 and $393,000 for fiscal year 2000 for the same activities.

TVA has also indicated that monitoring of impacts of the possible utility
deregulation scenarios has been a continuing process for the past several
years, and new aspects present themselves each year. TVA also said that it
had invested more than $1 million in the development of a new integrated
river system-modeling tool, called RiverWare, currently in use for daily
operations scheduling. TVA has been examining the possibilities of using
this model in a planning mode for future studies, as a complementary tool
to the Weekly Scheduling Model, or possibly as a replacement. TVA’s
annual efforts under the Flood Risk Reduction Program in recent years
have focused on the development of flood damage curves for various
communities within the Tennessee Valley. These curves depict the flood
damage that would occur at these sites as a function of the flood level on
the river. TVA has also developed flood distribution diagrams for 19 sites
throughout the Valley. These diagrams present a chart of the historical
flood events that occurred at these sites over the period of record. Flood
profiles have also been established for several lakes, which estimate the
elevations along the lakes and the statistical frequency at which these
elevations occur. TVA spent $551,000 during fiscal year 1996 and $849,000
during fiscal year 1997 on flood risk reduction activities. TVA’s budget
allocations for fiscal years 1998 and 1999 provide $500,000 each year to
continue the development of flood risk indicators and profiles and flood
damage curves. TVA has increased its budget allocation for these activities
to $520,000 for fiscal year 2000. In addition, TVA has cosponsored an effort
conducted by the National Weather Service to update and improve its
rainfall frequency analysis for the Tennessee Valley. TVA’s portion of the
funding totaled $250,000 during fiscal years 1997 and 1998.

TVA has stated the overall reasons why the 4-year moratorium should
continue. However, TVA has not informed the public and other
stakeholders of its recent and planned activities regarding a future
reevaluation of its policies impacting lake levels. By conducting internal
studies and developing the necessary tools and methodologies before


Page 61                         GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 5
Formal and Continuing Communication
Process With the Public and Other
Stakeholders Needed in Future Lake-Level
Study




announcing a comprehensive reevaluation of TVA’s policies impacting lake
levels, the task force believes TVA would reduce the risk of raising public
expectations of “immediate” changes in operating policies and would
help ensure that the significant resources required for a NEPA study would
be well spent. The task force recognized that TVA needed to develop and
implement a “proactive communication plan to increase public
understanding of the operation of its system of projects, reduce conflict,
and build rapport with stakeholders.” The task force recommended that
TVA incorporate public participation in its studies, where appropriate, to
ensure the credibility of the studies.

The July 1998 report recognized that the purpose of the communication
plan is to “better explain lake operating policies and tradeoffs to Members
of Congress, their staffs, distributors, lake users, and other stakeholder
groups” and to demonstrate TVA’s “continued willingness to listen to and
better understand stakeholder concerns.” The report also noted that two
of the key messages that TVA needed to communicate were (1) any
lake-level changes will shift benefits with impacts on flood risk, power,
navigation, and conditions for aquatic life and that evaluating these
impacts is a complex task requiring time, money, and broad input, and
(2) TVA’s commitment to improving communication with stakeholder
groups. In addition, the report indicated that several new mechanisms to
increase public involvement in river system operations would be
evaluated. One example was annual or semiannual workshops for lake
user groups. These workshops would include presentations on operations
of its system of projects and associated tradeoffs, opportunities for
stakeholders to bring concerns to the attention of TVA management, and
interaction among specific interest groups, such as recreation, flood
reduction, and navigation beneficiaries. Implementing the communication
plans outlined in TVA’s task force report would help to ensure that the
public and other stakeholders are kept informed of TVA’s activities and
future plans as TVA prepares for a reexamination of its policies impacting
lake levels.




Page 62                         GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                          Chapter 5
                          Formal and Continuing Communication
                          Process With the Public and Other
                          Stakeholders Needed in Future Lake-Level
                          Study




                          When TVA conducts its reevaluation of its policies impacting lake levels,
TVA Has Several           one of the more significant concerns it must consider is the effect of any
Options Available         policy changes on the environment of the Tennessee Valley. The National
When Assessing the        Environmental Policy Act of 1969 requires federal agencies to prepare a
                          detailed statement on the environmental impact of proposed major federal
Environmental Effects     actions that could significantly affect the quality of the human
of Alternative Policies   environment. In conducting its 1990 study, TVA developed an
                          environmental impact statement (EIS) to ensure that the environmental
Impacting Lake Levels     effects of the operating alternatives were thoroughly investigated and that
                          ample opportunities for public review and comment were provided.
                          However, under NEPA, there are several different levels of review that can
                          be used to analyze the environmental effects of policy changes.

                          Under the NEPA regulations, there are three levels of NEPA review:
                          environmental assessments (EA), EISs, and categorical exclusions. EAs are
                          supposed to be and typically are brief documents that still provide some
                          level of detail about proposed actions and alternatives, may have some
                          public involvement, and typically take several weeks to a number of
                          months to complete. At the end of an EA process, the agency either issues a
                          Finding of No Significant Impact or initiates an EIS process. EISs provide
                          the most detailed analyses of a proposed action and alternatives, have the
                          most formal public involvement process, and typically take the longest to
                          complete. Categorical exclusions are basically categories of actions that
                          an agency predetermines would normally not result in significant
                          environmental impacts. Under NEPA regulations, agencies can proceed with
                          an action that qualifies as a categorical exclusion without conducting any
                          further environmental review.

                          On the other hand, courts have recognized that NEPA is not applicable to
                          the continued operation of projects when operations conform to statutory
                          directives or when operations do not deviate from existing design
                          capacities and constraints. In addition, decisions about water flow through
                          federally operated dams and related matters have consistently been held
                          not to constitute major federal actions. Instead, courts have considered
                          such activity as “routine managerial actions” not subject to NEPA.

                          In the final analysis, however, if TVA believes it to be necessary or
                          advisable to prepare an EIS or perform any other type of environmental
                          review under NEPA, that decision would likely be respected by the courts.
                          Our review found no federal court decisions in which preparing an EIS was
                          found to have been unreasonable. NEPA was intended to facilitate reasoned




                          Page 63                         GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
              Chapter 5
              Formal and Continuing Communication
              Process With the Public and Other
              Stakeholders Needed in Future Lake-Level
              Study




              and informed decision-making, in which the environmental impacts of
              federal actions would be given due weight.

              According to TVA officials, TVA approaches NEPA from both a policy and a
              legal perspective. TVA recognizes that a wide range of operational changes
              can be made without evoking the NEPA process. However, TVA believes that
              there can be proposed changes in operations that could result in
              environmental changes that should be understood before decisions are
              made. By adopting the three levels of review created by NEPA, TVA does
              conduct a limited environmental review of most of the actions that are
              proposed to be categorically excluded and documents this review in
              writing. TVA believes that because it operates its projects as an integrated
              system and balances multiple purposes, any changes to its operating
              policies should be examined to determine the effects on the environment.
              However, a complete EIS is not necessary in all instances. Assuming the
              proposed change is not categorically excluded, TVA officials said that it is
              likely that TVA’s actions would involve at least the preparation of an EA or,
              if substantial enough of a change, a supplement to the existing EIS or
              another EIS. According to these officials, TVA anticipates that considerable
              updating of the material in the 1990 review would be required to more fully
              explain and document all environmental changes that might result from
              the implementation of any other alternatives.


              It has been nearly a decade since the last significant changes to TVA’s
Conclusions   policies on lake levels. During the current moratorium, TVA has initiated
              various internal actions through its lake level policy task force in
              preparation for a future reexamination of existing policies. TVA has been
              hesitant to make a formal announcement that a reexamination is needed
              because (1) it does not have all of the needed evaluation methodologies in
              place to perform a reexamination and (2) such an announcement would, in
              TVA’s view, set unrealistic expectations about how quickly any decisions
              could be reached about whether changes are or are not needed, given the
              time needed to conduct the reevaluation under the NEPA process.

              We agree with TVA that a reexamination of its policies impacting lake levels
              is warranted. An important aspect for TVA to consider in its reexamination
              efforts is formal and continuing communication with the public and other
              stakeholders. These communications are needed to (1) further educate TVA
              regarding the concerns and needs of the various stakeholders that must be
              considered in the reexamination process, (2) give TVA additional
              opportunities to explain the operation of its integrated system and the



              Page 64                         GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                     Chapter 5
                     Formal and Continuing Communication
                     Process With the Public and Other
                     Stakeholders Needed in Future Lake-Level
                     Study




                     complexities involved in evaluating changes to the system, (3) establish
                     realistic expectations of the time required to reevaluate changes in policies
                     impacting lake levels, (4) keep the public informed of TVA’s ongoing
                     activities and progress achieved, and (5) increase the overall credibility of
                     the reexamination process.

                     Past evaluations examining changes to policies impacting lake levels have
                     tended to emphasize either the costs associated with the potential change
                     as has been the case with TVA’s efforts or the localized benefits as has been
                     the case with studies performed for users of the lakes. When reexamining
                     any potential changes to policies impacting lake levels, a balanced and
                     comprehensive decision can only be reached through consideration of the
                     costs and benefits of the alternatives examined.


                     We recommend that the Chairman of TVA’s Board of Directors (1) provide
Recommendation       for a formal and continuing communication process for the public and
                     other stakeholders to actively participate in TVA’s efforts to reexamine its
                     policies impacting lake levels and (2) ensure that TVA’s reexamination
                     efforts include consideration of both the costs and benefits of any
                     potential changes to policies impacting lake levels.


                     TVA stated that we had conducted a comprehensive assessment of TVA’s
Agency Comments      tributary lake operating policies and that the draft report fairly
and Our Evaluation   summarized the issues influencing TVA’s operations. Regarding our
                     recommendation that TVA provide for a formal and continuing
                     communication process, TVA stated that it is essential that the public
                     continue to be involved in decisions that affect how the Valley’s water
                     resources are used. TVA also recognized that there is an opportunity to
                     improve its communications with stakeholders. TVA added that it remains
                     committed to communicating fully with its stakeholders and others who
                     depend on the integrated management of the Tennessee River system.

                     Regarding our recommendation that TVA’s reexamination efforts consider
                     the costs and benefits of potential changes, TVA stated that in operating the
                     river system for the greatest public benefit, it continues to look for new
                     methodologies that will provide for a more objective analysis of the
                     tradeoffs among competing demands. TVA stated that certain water uses,
                     such as hydropower and flood control, lend themselves more readily to
                     quantitative analysis, while other operating objectives, such as economic
                     development and environmental impacts, continue to be more difficult to



                     Page 65                         GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Chapter 5
Formal and Continuing Communication
Process With the Public and Other
Stakeholders Needed in Future Lake-Level
Study




quantify. TVA added that it remains committed to providing lake users and
other beneficiaries with the best information it has on the most likely
impacts of changes in lake operations so that such lake users and other
beneficiaries are aware of the tradeoffs and consequences of policy
changes.

TVA also emphasized in its comments that not all costs are monetary. TVA
stated that as it works to meet multiple needs with a finite resource,
increased costs can take the form not only of higher electricity prices but
also reduced flood control benefits, lessened environmental quality, and
reductions in other benefits. TVA added that costs in any form become
costs to some segment of the public.




Page 66                         GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Page 67   GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix I

Comments From the Tennessee Valley
Authority




              Page 68   GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix I
Comments From the Tennessee Valley
Authority




Page 69                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix II

Multipurpose Tributary Lake-Level
Changes—Minimum and Maximum Monthly
Levels During Calendar Year 1998

In feet above sea level
Months during
1998                      Blue Ridgea    Boone     Chatuge    Cherokee        Douglas       Fontana      Hiwassee
January
  Minimum                    1,651.18   1,351.35   1,912.19     1,027.04        940.51      1,627.98      1,462.78
  Maximum                    1,664.27   1,359.19   1,916.05     1,033.25        961.37      1,655.43      1,485.81
February
  Minimum                    1,664.10   1,353.59   1,913.42     1,033.06        950.66      1,651.42      1,475.43
  Maximum                    1,678.13   1,361.90   1,915.60     1,036.82        960.99      1,656.09      1,481.55
March
  Minimum                    1,678.11   1,359.27   1,915.26     1,036.79        955.25      1,653.65      1,478.79
  Maximum                    1,684.30   1,365.55   1,919.59     1,051.83        971.45      1,660.21      1,491.88
April
  Minimum                    1,682.30   1,364.29   1,919.40     1,051.41        971.45      1,659.34      1,491.88
  Maximum                    1,689.07   1,379.38   1,924.55     1,069.27        995.36      1,693.52      1,515.95
May
  Minimum                    1,683.77   1,376.69   1,924.26     1,066.91        991.84      1,693.52      1,515.83
  Maximum                    1,687.18   1,383.80   1,925.35     1,070.99        993.98      1,702.75      1,524.25
June
  Minimum                    1,685.90   1,380.69   1,925.00     1,069.02        993.05      1,700.59      1,520.80
  Maximum                    1,687.95   1,383.24   1,926.31     1,071.76        995.20      1,705.34      1,524.30
July
  Minimum                    1,682.63   1,381.51   1,923.12     1,060.53        991.06      1,693.64      1,516.89
  Maximum                    1,686.09   1,383.72   1,925.15     1,069.31        993.23      1,700.65      1,521.08
August
  Minimum                    1,673.89   1,380.92   1,919.16     1,052.02        976.90      1,679.02      1,506.15
  Maximum                    1,682.68   1,383.04   1,923.12     1,060.75        991.06      1,693.69      1,517.40
September
  Minimum                    1,654.10   1,378.90   1,915.63     1,042.61        965.81      1,655.21      1,491.17
  Maximum                    1,673.96   1,382.50   1,919.16     1,052.09        976.91      1,679.02      1,506.38
October
  Minimum                    1,630.60   1,372.91   1,913.69     1,036.16        954.22      1,639.47      1,472.30
  Maximum                    1,654.10   1,379.00   1,915.80     1,042.67        966.04      1,655.25      1,491.25
November
  Minimum                    1,621.38   1,361.33   1,911.95     1,029.17        946.72      1,633.45      1,464.85
  Maximum                    1,630.65   1,373.12   1,913.69     1,036.76        954.33      1,639.60      1,472.59
December
  Minimum                    1,624.72   1,355.28   1,911.08     1,026.37        940.29      1,625.68      1,460.31
  Maximum                    1,636.16   1,361.63   1,912.05     1,030.07        946.78      1,633.45      1,465.42




                                        Page 70               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                         Appendix II
                         Multipurpose Tributary Lake-Level
                         Changes—Minimum and Maximum Monthly
                         Levels During Calendar Year 1998




Melton Hill    Norris        Nottely   South Holston                Tellico     Tims Forda          Watauga


791.46         977.02       1,742.39         1,689.40               807.51           856.01         1,930.63
794.88         986.00       1,749.67         1,699.26               810.60           866.16         1,940.20


791.17         986.00       1,748.28         1,699.26               807.45           865.01         1,940.01
794.98         992.63       1,754.07         1,712.01               810.50           874.00         1,948.68


790.10         992.63       1,753.82         1,712.01               807.51           874.00         1,948.02
794.99        1,006.52      1,761.98         1,723.05               810.60           881.94         1,956.42


790.79        1,006.43      1,761.98         1,721.65               808.25           881.69         1,954.13
796.44        1,030.38      1,772.47         1,732.29               815.09           886.25         1,960.69


792.05        1,020.15      1,772.34         1,728.43               811.72           884.60         1,956.83
794.97        1,024.25      1,775.98         1,731.84               813.11           885.98         1,959.42


792.30        1,019.15      1,775.39         1,728.46               811.88           885.84         1,957.24
795.05        1,021.23      1,778.09         1,730.64               814.42           890.14         1,959.85


790.40        1,015.00      1,770.29         1,721.30               812.07           886.92         1,949.22
794.88        1,019.15      1,775.53         1,728.90               813.30           887.68         1,957.88


792.45        1,006.96      1,762.82         1,713.28               812.17           886.18         1,942.60
795.10        1,015.03      1,770.41         1,721.32               813.48           888.06         1,949.30


792.58         996.87       1,755.19         1,703.21               812.37           883.68         1,935.52
794.71        1,006.99      1,762.85         1,713.28               813.30           886.19         1,942.62


790.27         987.52       1,747.59         1,697.18               812.05           880.66         1,934.03
794.89         996.88       1,755.28         1,703.26               813.13           883.93         1,935.68


792.68         980.95       1,744.95         1,696.31               808.94           875.61         1,933.80
794.63         987.55       1,747.64         1,697.25               812.69           880.68         1,934.11


792.48         977.39       1,740.84         1,693.65               807.78           870.88         1,932.17
794.98         981.18       1,745.01         1,696.34               809.61           875.61         1,934.04




                         Page 71                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix II
Multipurpose Tributary Lake-Level
Changes—Minimum and Maximum Monthly
Levels During Calendar Year 1998




a
 TVA said that the elevations for Blue Ridge beginning in September were abnormally low due to
a special drawdown for maintenance inspection that is conducted once every 5 years. TVA also
said that the elevations for Tims Ford Lake in January and February were abnormally low due to a
special drawdown for repairs.

Source: TVA.




Page 72                              GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix III

TVA’s 1990 Review and Its 1999 Analysis


                             This appendix provides information from TVA’s 1990 review, entitled
                             Tennessee River and Reservoir System Operation and Planning Review,
                             about the increased systemwide costs of supplying electric power that TVA
                             estimated for a range of lake-level alternatives and a description of the
                             methodologies used in TVA’s analysis. In addition, this appendix
                             summarizes the results of an analysis TVA performed in 1999 at our request
                             that updated its estimated systemwide cost impacts of implementing
                             lake-level alternatives similar to two of the alternatives considered in the
                             1990 review.



TVA’s 1990 Review

TVA Estimated                In 1990, TVA estimated that various alternative policies for managing lake
Systemwide Costs of          levels at some of its multipurpose tributary projects would result in
Supplying Electric Power     increasing its annual average systemwide costs of supplying electric
                             power by $2 million to $93 million (in 1990 dollars), depending on the
for a Wide Range of          alternative (see table III.1). TVA used a complex methodology to estimate
Lake-Level Alternatives in   the cost impacts of the seven lake-level alternatives, each of which
1990                         involved a delay of the starting date for the drawdown of lake levels from
                             Memorial Day to later dates in the summer or fall.

                             TVA’s methodology focused on characterizing the electricity supply
                             resources (commonly referred to as capacity) it would need and how it
                             would deploy these resources in order to meet demand from its customers
                             under the different lake-level alternatives. TVA computed differences in
                             resource needs between each alternative and the “base case” scenario.
                             The base case refers to a “no-change” scenario that maintains the existing
                             policy of starting the lake-level drawdown anytime after Memorial Day.
                             Differences in capacity needs and in the utilization and scheduling of
                             different sources of generation were then translated into cost differences.

                             For example, delaying the drawdown of the lake levels until Labor Day
                             (Alternative 2) was more expensive than the base case because less water
                             would be available for hydroelectric generation during the period of peak
                             electricity demand in late summer. The cost impact can be broken into




                             Page 73                     GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
    Appendix III
    TVA’s 1990 Review and Its 1999 Analysis




    two components, “energy costs” and “capacity costs.”1 For example,
    under Alternative 2:

•   Delaying the drawdown of the lake levels from Memorial Day to Labor Day
    meant that there would be less hydroelectric generation during a period of
    peak summer demand. The shortfall would be compensated by
    incremental production of electricity from other sources, such as
    coal-fired plants or combustion turbine units. However, the energy costs
    of fossil-fuel sources (coal, petroleum based, and gas) are higher than the
    energy costs of hydroelectric power generation. This pattern would be
    reversed in the fall and early winter, when more hydroelectric generation
    is available. However, because more electricity is demanded and produced
    in late summer than in early winter, the net effect would be higher energy
    costs.
•   The reduction of hydroelectric generation capacity due to delaying the
    drawdown of the lake levels to Labor Day was significant enough that TVA
    would have to acquire incremental fossil-fueled electric generation
    capacity to replace it. In effect, this meant that TVA could not rely on
    increasing the utilization of its existing plants, but it would have to build
    new capacity. This cost is over and above the higher energy costs.

    Table III.1 displays TVA’s estimated energy and capacity costs associated
    with each of the alternatives considered in the 1990 review.




    1
     Energy represents the volume of electricity delivered over a period of time and is measured in
    kilowatt-hours. Capacity, on the other hand, is measured in kilowatts and represents the rate at which
    energy can be delivered at a moment in time. For example, 20 100-watt light bulbs burning for one-half
    hour would use 1,000 watt-hours (or 1 kilowatt-hour). In this case, the capacity demanded would be 2
    kilowatts and the energy used would be 1 kilowatt-hour. Energy costs include fuel and other costs that
    depend on operating levels, but exclude capital costs. Capacity costs are the fixed costs of an electric
    power system (such as the capital cost of a generation plant) that do not vary with the level of
    operation.



    Page 74                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                          Appendix III
                                          TVA’s 1990 Review and Its 1999 Analysis




Table III.1: Estimated Impact of Lake-Level Alternatives Evaluated by TVA in 1990 on TVA’s Systemwide Cost of Supplying
Electric Power
In millions of 1990 dollars
                                                      Required capacity
                       Annual energy cost                   addition (in                                    Annual capital         Total annual
Alternative             Average            Range            megawatts) Initial capital cost                          cost                  cost
1                            $2       $ –9 to $20                          0                       0                        0                  $2
1A                           $6       $ –7 to $ 30                      100                     $ 74                      $9                 $ 15
1B                           $9       $ –9 to $ 20                        10                     $7                       $1                 $ 10
1C                           $3       $ –9 to $ 20                        30                    $ 22                      $3                   $6
1D                           $3       $ –9 to $ 20                         0                       0                        0                  $3
2                           $ 16      $ –4 to $ 60                      750                   $ 560                     $ 68                 $ 84
3                           $ 25      $ –4 to $ 56                      750                   $ 560                     $ 68                 $ 93
                                          Source: TVA, Tennessee River and Reservoir System Operation and Planning Review, Table 28,
                                          p. 116, (Dec. 1990).




TVA’s 1990 Cost Estimates                 TVA’s 1990 cost-estimation methodology involved the use of three
Are Based on Complex                      computer models and extensive data inputs to simulate its hydroelectric
Modeling and Extensive                    generation capabilities and operations, and its overall costs of supplying
                                          electric power using all its electricity generation assets under the different
Data                                      alternatives. The modeling effort accounted for considerable uncertainty
                                          in hydrologic conditions, which leads, in turn, to uncertainties in cost
                                          estimates.

                                          To determine differences in the cost of electric supplies between
                                          alternative lake policy scenarios and the base case (no-change) scenario,
                                          TVA used three models. TVA also used extensive data on hydrological
                                          conditions; its own electricity supply capabilities, operations and costs;
                                          and its customers’ electricity demand. On the basis of the data, the models,
                                          and standard analytical techniques, TVA determined (1) its hydro and
                                          nonhydro electricity generation capabilities under each lake-level
                                          alternative, (2) the most efficient way it could schedule its power plants to
                                          meet demand, and (3) its costs of meeting demand under each alternative.
                                          TVA conducted this analysis for 1 year, 1993, when electric supplies from
                                          its own sources were expected to equal demand.2 TVA examined
                                          hydrological uncertainty by applying the one year of incremental power
                                          supply costs for 1993 to each of the previous 87 years (1903-1989) for

                                          2
                                           According to TVA, it chose a balanced year, 1993 (in which its self-generated supply of electricity was
                                          expected to equal demand), to avoid overestimating or underestimating the cost of delaying the
                                          drawdown of lake levels.



                                          Page 75                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix III
TVA’s 1990 Review and Its 1999 Analysis




which it had hydrological data. Briefly, TVA’s modeling procedure was as
follows:

The Weekly Scheduling Model (WSM): This is a TVA-developed model that is
used to simulate weekly hydroelectric power capabilities and production
based on hydrologic conditions. WSM considers various constraints on
water releases imposed to meet navigation, flood control, environmental,
and recreational objectives. TVA used WSM to determine weekly
hydroelectric production under each lake-level scenario.

PowrSym Dispatch Model: This is a power production costing model that
is used to determine the most efficient combination of electric generation
assets to meet a given level of demand. This model uses data on TVA’s
hydro and nonhydro power generation to determine which plants should
generate the needed electricity to meet demand at any given time, and it
calculates associated energy costs. For the lake-level scenario analysis, TVA
fed the weekly hydroelectric power production data into the PowrSym
model to calculate its energy costs under each lake-level scenario.
PowrSym scheduled the weekly hydropower production on an hourly
basis together with all of TVA’s other power resources to minimize overall
production costs.

The Hourly Loss of Load Expectation (HLOLE) Model: This model is used to
determine electric generation capacity needed to maintain the reliability of
the power system. TVA input the power supply data resulting from the
previous steps into the HLOLE model in order to determine the different
electric generation capacity needs under each alternative and to compare
them to the capacity needs of the base case. On the basis of assumptions
on the construction costs of certain types of power plants, TVA then
converted the resulting differences to capacity cost differences among
scenarios.3 The hydrologic uncertainties as depicted in 87 years of highly
variable weather-related data, resulted in wide ranges of cost estimates.4
For example, the annual average estimate of the energy cost of Alternative
2, as reported in table III.1, was $16 million (in 1990 dollars). However, TVA
reported a wide range of uncertainty around this average.5 When required

3
 In 1990, TVA assumed that any expansion of its power generation capacity would be in the form of
new coal-fired and combustion turbine power plants. A combustion turbine is an electricity generating
unit that is similar in design to a jet engine. The combustion turbine uses natural gas or other light
hydrocarbon fuel to fire a turbine engine that is connected to an electric generator.
4
 In addition to considering an average-year for 87 years of hydrology, TVA considered hydrologic
scenarios based on an extremely dry year and an extremely wet year among the 87 years.
5
 TVA reported a range between $(–)4 million (negative $4 million, that is, benefits instead of costs)
annually, and up to $60 million annual energy costs (in 1990 dollars).



Page 76                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                              Appendix III
                              TVA’s 1990 Review and Its 1999 Analysis




                              capacity additions were included, the total average annual cost was
                              estimated at $84 million (in 1990 dollars).



TVA’s 1999 Analysis

TVA Updated Systemwide        We asked TVA to update its cost estimates of two lake-level alternatives
Costs of Supplying Electric   because market conditions and evaluation methodologies have changed
Power for Two Lake-Level      considerably since 1990. One key difference in the 1999 update was that
                              the 1990 base case was no longer the base case in 1999. In 1991, TVA
Alternatives in 1999          decided to adopt Alternative 1 as a permanent policy. This decision meant
                              delaying the start of the lake-level drawdown at TVA’s multipurpose
                              tributary projects from Memorial Day to August 1. As a result, Alternative
                              1 became the new base case. Furthermore, the electricity supply industry
                              had changed considerably, due to deregulation and the introduction of
                              competition;6 technological progress; and environmental factors. We and
                              TVA agreed on the broad issues to be considered in TVA’s cost updates. Due
                              to the difficulty of trying to update all of the alternatives considered in the
                              1990 analysis, we requested that TVA update only two
                              alternatives—Alternative 1A (an additional 2-month drawdown delay for
                              three of the projects) and Alternative 2 (an additional 1-month drawdown
                              delay for all of the projects). The results of TVA’s March 1999 analysis, as
                              detailed in table III.2, indicate average annual cost increases of $14 million
                              and $47 million (in 1999 dollars) in its systemwide cost of supplying
                              electric power for Alternative 1A and Alternative 2, respectively, but with
                              very wide ranges of uncertainty for each.




                              6
                               In the past, electricity was supplied by regulated entities (utilities) that acted as monopolies within
                              their franchise areas. Monopolistic utilities generated their own electricity supplies and sold them
                              through their distribution systems and were allowed to charge rates based on their costs plus a
                              regulated rate of return. As deregulation progresses, electricity supplies are being increasingly
                              produced by independent companies operating in unregulated, competitive markets. The current trend
                              is for the monopolistic function of utilities to be increasingly limited to their transmission and
                              distribution systems. Less and less of the electricity is self-generated by utilities and more by
                              unregulated suppliers. The utilities purchase the electricity from the unregulated suppliers at
                              competitive rates and use its transmission and distribution system to transport it to customers in the
                              franchise area.



                              Page 77                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                           Appendix III
                                           TVA’s 1990 Review and Its 1999 Analysis




Table III.2: Estimated Impact of
Lake-Level Alternatives Evaluated by       In millions of 1999 dollars
TVA in 1999 on TVA’s Systemwide                                                         Annual                   Annual                   Annual
Cost of Supplying Electric                 Alternative                             average cost                cost—low                cost—high
Power—Differences Between
                                           1A                                                 $14                    $(-) 2                     $33
Alternatives and Base Case
                                           2                                                  $47                         0                     $88
                                           Source: TVA.



                                           TVA’s procedures for updating cost estimates in 1999 differ in several
                                           significant ways from its 1990 estimation procedure, largely because
                                           conditions have changed. The differences include the following:

                                       •   The 1999 base case is different, as mentioned earlier.
                                       •   Whereas the 1990 procedure estimated annual costs for a single year, 1993,
                                           the 1999 procedure estimated average annual costs for the entire future
                                           period, 1999 through 2023. TVA broke this 25-year period into 2 sub-periods,
                                           1999 through 2001, and 2002 through 2023.
                                       •   The 1999 methodology uses the WSM model in a similar fashion to
                                           determine different hydroelectric capacities and production for each of
                                           the scenarios. However, the 1999 methodology uses a 25-year future time
                                           horizon for the analysis and breaks the 25 years into two periods. During
                                           the first 3 years, TVA would purchase forward contracts from electric
                                           power suppliers outside its territory in order to replace lost hydropower,
                                           while additional generating capacity would be built to compensate for its
                                           reduced hydro generation capacity.7 The transmission costs associated
                                           with importing this power into the TVA power system were included in the
                                           cost estimates for these 3 years.8 TVA assumes that, starting in 2002, the
                                           newly built capacity will supply any shortages due to changes in policies
                                           impacting lake levels. However, in contrast with its 1990 approach, TVA did
                                           not calculate the cost of lake-level alternatives based on its own costs of
                                           the energy and capacity needed to compensate for reduced hydroelectric
                                           power generation. Instead, TVA used its forecast market prices in the cost
                                           calculations.9 This change in approach reflects the trend away from


                                           7
                                            A forward contract is an agreement between two parties for the purchase/sale of electric supplies at
                                           some future time under such conditions as the two agree on. The forward contracts that TVA
                                           purchases are for firm power, that is, power that the seller guarantees will be delivered to the buyer.
                                           Firm power is priced in a way that reflects both energy and capacity costs.
                                           8
                                            Transmission is the process of conducting the flow of electricity at high voltages from the points of
                                           generation to the location of groups of electricity users. TVA assumed that it would obtain
                                           transmission at a price of $1.80/kilowatt-month for the 3-month summer period (July-September).
                                           9
                                            TVA actually used three sets of forecast market prices representing a low-, medium-, and high-price
                                           scenarios.



                                           Page 78                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
    Appendix III
    TVA’s 1990 Review and Its 1999 Analysis




    regulated cost-based prices and towards competitive market-based prices
    in the electric industry. The main difference between the two approaches
    is that the 1990 cost-based approach assigns all the costs of the additional
    capacity to each lake-level alternative. In contrast, the use of market
    prices in the 1999 analysis means that only a portion of the cost of
    additional capacity is assigned to each lake-level alternative.10
•   The use of TVA’s models in 1999 was different from 1990. The key
    difference is that PowrSym was used in 1990 to model only TVA’s power
    system in order to estimate energy cost differences among scenarios. In
    1999, however, the PowrSym Multi-Area model (which is different from
    the PowrSym production costing model) was used to forecast electricity
    market prices for power in 23 southern, midwestern, and some eastern
    states.
•   During the 3-year period, 1999-2001, TVA’s analysis does indicate the need
    for additional generation capacity to compensate for lower hydroelectric
    capacity under Alternatives 1A and 2. However, TVA’s analysis of delays in
    the drawdown of lake levels yielded smaller impacts on hydroelectric
    generation capacity in 1999 than in 1990. For example, in 1990 TVA
    estimated that Alternative 2 would require TVA to build 750 megawatts
    (MW) of additional electric generation capacity. In contrast, in 1999, TVA
    estimated that Alternative 2 would require TVA to build 250 MW of
    additional capacity. This difference is due to the different base case
    (August 1st start date of the lake-level drawdown instead of Memorial
    Day), technological advances resulting in lower construction costs and
    higher power plant efficiencies for new capacity, improved reliability of
    TVA’s nuclear and fossil-fueled plants, and improved availability of power
    for purchase from other suppliers. For the 1999 analysis, the type and cost
    of electric generation capacity that TVA acquires during the first 3 years
    differ considerably from its assumed capacity additions in the 1990
    methodology. This reflects technological progress, increased efficiency,
    and lower costs.11



    10
      It should be noted that the additional capacity may or may not be built by TVA. It may be built by
    another power producer that sells the power to TVA at market rates. Whether it is built by TVA or by
    another producer, the additional capacity will supply power to TVA’s customers only during the peak
    hours affected by the change to policies impacting lake levels. Outside these peak hours, the power
    from the additional capacity can be sold elsewhere at market rates. Since market rates effectively
    reflect both energy and capacity (capital) costs, not all the capital costs of the replacement capacity
    are counted towards the cost of the lake-level alternative. The remaining capital costs are paid by
    other users of this capacity who purchase power outside of these peak hours.
    11
     TVA assumes the expansions of its electric generating capacity during the 3-year construction period
    will be in the form of a combined cycle plant. Combined-cycle power plants are generating units that
    combine a combustion turbine and a heat recovery steam generator. The steam generator uses the
    exhaust from the combustion turbine to generate steam that, in turn, drives a steam turbine.



    Page 79                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
    Appendix III
    TVA’s 1990 Review and Its 1999 Analysis




•   The 1990 analysis did not consider the impacts of lake-level alternatives on
    cooling requirements for some of its fossil-fuel generating plants. The 1999
    analysis added some costs associated with such impacts.12
•   The 1999 analysis also assigns a value for the “ancillary services”
    associated with the flexibility of hydroelectric generation.13 This value
    raised TVA’s estimates of the impacts on its systemwide cost of supplying
    electric power of the lake-level alternatives by 10 to 15 percent—a range of
    0 to $4 million for Alternative 1A and from 0 to $11 million for Alternative
    2 (in 1999 dollars). This change in the 1999 methodology is due to the
    assumption that market-like transactions for ancillary services have
    emerged and are likely to grow.14
•   TVA handled hydrologic and electricity price uncertainties a different way
    for the 1999 analysis, reflecting the fact that the 1999 analysis had a
    25-year future time horizon, while the 1990 analysis focused on a single
    year, 1993.15
•   TVA’s 1999 analysis also considered the effect of more stringent future air
    pollution regulations to take effect in 2010, which would essentially make
    hydropower more valuable. TVA’s 1999 analysis, using the more stringent
    regulations, would result in somewhat higher estimated impacts.16 This
    was not a factor in the 1990 analysis.

    12
     TVA told us that changes to policies impacting lake levels have implications on water use to cool
    some of its nuclear and fossil-fired electric plants. Given water temperature requirements in National
    Pollution Discharge Elimination System permits, changes in policies impacting lake levels could
    effectively reduce the output of some plants. TVA estimated the costs related to cooling requirements
    as ranging from 0 for Alternative 1A to $1 million to $4 million for Alternative 2 (in 1999 dollars).
    13
      Ancillary services are defined as those services that are necessary to support the transportation of
    power from power plants to customers while maintaining reliable operation of the transportation
    system in accordance with good utility practice. Hydroelectricity commands a premium for its
    “ancillary” value, associated with its operational versatility that is particularly valuable for regulating
    the transportation of electricity on power lines. The versatility of hydroelectric power plants lies in the
    fact that they can be brought into service quickly and cheaply. In contrast, bringing a large coal-fired
    power plant into service and taking it off-line is both time-consuming and costly.
    14
      TVA told us that there is considerable uncertainty regarding how ancillary services will be valued in
    future electricity markets. TVA believes that the use of a 10-percent to 15-percent cost escalation
    factor was conservative, based on recent evidence of the valuation of hydroelectric power’s versatility.
    15
      For the 1999 analysis, TVA used 96 years of hydrologic data (1903-1998) and a 25-year forecast of
    hourly electricity prices. For each lake-level alternative, WSM processed the 96 years of hydrology into
    72 sets of simulations, each consisting of 25 years of hydropower system performance data, which
    were multiplied by the 25-years of forecasted hourly electricity prices to estimate costs. The first set
    started with 1903 and ended with 1927; the second set started with 1904 and ended with 1928; etc.,
    until the 72nd set, started with 1974 and ended with 1998. The 72 sets differed in terms of hydroelectric
    generation, depending on the hydrologic data of each of them. TVA reported its cost estimates to
    reflect a range from particularly dry, to medium, to very “wet” sets of 25 years. As mentioned, the
    price uncertainty was addressed by using three sets of forecast market prices.
    16
     For example, TVA estimated that these more stringent regulations would increase the average
    systemwide cost to supply electric power for Alternative 1A by $1.2 million and for Alternative 2 by
    $2.3 million (in 1999 dollars and based on a medium price forecast).



    Page 80                                 GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                 Appendix III
                                 TVA’s 1990 Review and Its 1999 Analysis




                             •   TVA’s 1990 analysis relied on a discount rate of 7 percent. In the 1999
                                 analysis, TVA used both a 6-and an 8-percent discount rate. The 6-percent
                                 rate resulted in marginally higher costs than the 8-percent rate in the 1999
                                 analysis.
                             •   TVA’s 1999 analysis also recognized that a change in TVA’s policy impacting
                                 lake levels would also impact a class of direct industrial customers.17

                                 TVA officials told us that the 1999 estimates of cost impacts of the
                                 lake-level alternatives were lower than the respective 1990 estimates for a
                                 number of reasons. As mentioned earlier, one reason is that the base case
                                 in 1999 (August 1st lake-level drawdown) is different from the 1990 base
                                 case (Memorial Day lake-level drawdown). In the 1990 analysis, for
                                 example, Alternative 2 required TVA to acquire 750 MW of additional electric
                                 generation capacity, as compared to an estimated 250 MW for Alternative 2
                                 in the 1999 analysis. Furthermore, whereas the entire capital cost of the
                                 750 MW was counted in the cost estimate of delaying the lake-level
                                 drawdown in 1990, the same was not true for the cost of the 250 MW in
                                 1999. Electricity deregulation and the increasing trend to competitive
                                 market pricing contribute to lowering costs and improving the utilization
                                 of resources. Technological advances have lowered the cost of electricity
                                 and improved operating efficiencies.


TVA Emphasized That              TVA officials told us that great uncertainties in today’s electricity supply
Uncertainties Exist in Its       industry go well beyond the uncertainty that its 1999 modeling exercise
Cost Analyses                    attempted to depict. Electricity deregulation has started in some states
                                 and is being considered in other states and possibly at the federal level.
                                 TVA assumes that electricity deregulation will result in the development of
                                 an electricity spot market. TVA assumes that, starting in the year 2002, it
                                 can rely on such a market for power purchases to make up for any
                                 shortfalls that will result from alternative policies impacting lake levels.




                                 17
                                  These customers have Economy Surplus Power contracts with TVA, which allow them to buy a
                                 portion of their electricity needs at low rates (TVA’s hourly marginal costs) because TVA has the
                                 option to turn off their power supplies during periods of peak demand when it needs more supplies to
                                 serve its firm customers. Firm customers pay higher rates because they have a higher priority. Hourly
                                 marginal costs (and hence Economy Surplus Power rates) will increase during the summer because
                                 more expensive generation sources will be used to replace hydropower losses due to delays in the
                                 drawdown of lake levels.



                                 Page 81                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix III
TVA’s 1990 Review and Its 1999 Analysis




However, how quickly such a market will develop and how reliable it will
be remains to be seen.18

Environmental regulation may also have significant consequences on the
industry, possibly causing major shifts from coal-fired electricity
generation to less polluting sources of generation. International
agreements on global warming, resulting in even more stringent
environmental regulations, are an added source of uncertainty. All of these
factors may have profound impacts on electricity costs and prices in the
future.




18
  As mentioned, TVA’s 1999 analysis assumed that, during the period 1999-2001, it would rely on
forward contracts from electric power suppliers outside its region for power losses due to lake-level
alternatives. TVA officials, however, also said that this assumes that there will be sufficient
transmission capacity to import the needed power. TVA officials, however, think that transmission
capacity is another source of uncertainty that they did not model explicitly. Transmission capacity may
indeed constrain TVA’s ability to rely on power purchases from suppliers outside of its region.



Page 82                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix IV

Study Examining Cherokee and Douglas
Lakes

                        This appendix gives a brief description of the study entitled, Economic
                        and Fiscal Consequences of TVA’s Draw-Down of Cherokee and Douglas
                        Lakes (henceforth, the Cherokee and Douglas Study), which was prepared
                        by the University of Tennessee’s Center for Business and Economic
                        Research. As mentioned in chapter 4, this study used different approaches
                        to estimate the positive economic impacts of lake-level changes on areas
                        bordering the two lakes. This appendix also summarizes views of the study
                        expressed by representatives of the LOUD and TVA, and our comments on
                        the study.


                        LOUD and CLUA representatives believe that their proposed TVA changes to
Study Estimates Local   policies impacting lake levels will have positive economic and fiscal
Benefits Resulting      impacts on the six counties bordering the two TVA lakes. According to the
From Proposed TVA       study, LOUD’s and CLUA’s proposals for changes to policies impacting lake
                        levels are closest to a set of lake draw-down delays that TVA considered in
Changes to Policies     its 1990 review as Alternative 1A.1 The Cherokee and Douglas Study
Impacting Lake Levels   commissioned by these groups estimated that delaying the drawdown of
                        the lakes would increase regional income by $0.6 million to $5.7 million.

                        The study used different approaches to estimate the impact of lake-level
                        changes on income, employment, and local sales tax revenues of the six
                        counties surrounding the two lakes. One approach started with an
                        estimate of the impact of delayed drawdown on expenditures in the
                        six-county region due to increased visitation to the two lakes by
                        non-county residents. The researchers interviewed about 75 visitors who
                        came to Douglas and Cherokee from outside the six-county region.2 The
                        survey asked respondents about their expenditures within the six-county
                        region during their visits to the two lakes. Each respondent was shown a
                        picture of Douglas Lake at an elevation of about 20 feet below normal
                        summer recreational season level. Respondents were then asked if a delay
                        in drawdown would result in him or her making more or less visits to the
                        Douglas or Cherokee lakes and to other area lakes. On the basis of data


                        1
                         This alternative calls for delaying drawdown at Douglas and Cherokee, and a third lake, Norris, from
                        Memorial Day to October 1, in addition to delaying lake-level drawdown at seven other tributary
                        projects in eastern Tennessee to August 1. According to TVA, the LOUD and CLUA proposals are
                        different from Alternative 1A as defined in the 1990 EIS. The LOUD and CLUA proposals request
                        significant changes in the January 1 target lake levels, but Alternative 1A does not consider such
                        changes. The study did not independently estimate any costs resulting from the proposed changes. The
                        study reported TVA’s 1990 estimated impacts of the proposed changes on the systemwide cost of
                        supplying electric power, adjusted for inflation to reflect 1998 values.
                        2
                         The researchers conducted 161 in-person interviews of visitors to Douglas and Cherokee, with
                        roughly an equal number of visitors (80) being interviewed at each of the lakes. Of those interviewed,
                        the percentages of nonresidents were 59 percent at Douglas and 36 percent at Cherokee.



                        Page 83                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix IV
Study Examining Cherokee and Douglas
Lakes




gathered from the survey, the study estimated increases in retail
expenditures in the six counties due to increased visitation by
nonresidents of about $1 million to $1.8 million per year. (See table IV.1.)

Two other approaches in the study started by estimating the impact of
changes to policies impacting lake levels on total retail sales in the
six-county region.3 The first approach used statistical techniques to
analyze whether there was a significant relationship between monthly lake
levels and the region’s monthly retail sales. The analysis confirmed such a
relationship and estimated that drawdown delays from August 1 to
October 1 would add about $1.6 million to annual total retail sales in the
six-county region. The second approach based on retail sales impact was
based on 200 responses to a survey of commercial establishments in the
six-county region. The survey asked businesses to estimate how much
their retail sales were likely to increase if drawdown of the two lakes were
delayed. While a total of 1,088 surveys were sent to selected commercial
establishments in the six-county region, the response rate to the survey
was only about 18 percent. The results of this survey were reported in the
study as an estimated increase in annual retail sales of about $7.1 million
due to the drawdown delays.

The study used these estimates of expenditures and retail sales increases
to derive measures of impacts of changing lake levels on income,
employment, and sales tax revenues in the six-county region. The annual
impacts due to lake drawdown delays were estimated as follows: increases
in income of about $0.6 million to $5.7 million; increases in employment of
between 205 and 2,106 full-time job equivalents; and increases in local
sales tax revenues of about $39,000 to $239,000.4 (See table IV.1.)




3
 The study did not attempt to isolate the impact on retail purchases by nonresidents of the six-county
region. Instead, the impact of changes to policies impacting lake levels on total retail sales was
estimated.
4
 The study reported the results of two other approaches that the researchers used to estimate
drawdown delay benefits; namely, a “benefits transfer approach” and another approach based on
“angler’s choices of reservoirs.” The former yielded relatively low benefit estimates, while the latter
yielded a “conservative estimate of losses to anglers” of $11 million annually. These results, however,
were not reported in the study’s summary table entitled “Comparison of Economic Impacts Resulting
from Higher Lake Levels in August and September for Douglas and Cherokee Lakes.”



Page 84                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                        Appendix IV
                                        Study Examining Cherokee and Douglas
                                        Lakes




Table IV.1: Comparison of Measures of
Economic Impacts Resulting From                                                                                          Commercial
Higher Lake Levels in August and                                          Survey of lake       Statistical retail     establishments
September for the Douglas and           Economic measure                        visitors               analysis          retail survey
Cherokee Lakes                          Increased expenditures/               $1 million -
                                        retail sales                          $1.8 million           $1.6 million          $7.1 million
                                        Income impact                        $0.6 million -
                                                                                $1 million           $0.7 million          $5.7 million
                                        Employment impacta                      205 – 357                     259                 2,106
                                                                             full-time jobs        full-time jobs         full-time jobs
                                        Local sales tax revenue        $48,117 - $84,185                $38,868              $239,187
                                        a
                                        Full-time equivalent jobs; assumes all employment gains accrue in August and September.

                                        Source: Center for Business and Economic Research, The University of Tennessee, Economic
                                        and Fiscal Consequences of TVA’s Draw-Down of Cherokee and Douglas Lakes, Knoxville, TN, p.
                                        xii, (Oct. 1998).



                                        In addition to estimates of impacts on the six-county income, employment,
                                        and sales tax revenues, the study pointed out that delays in the drawdown
                                        of lake levels would likely have significant impacts on the values of
                                        properties that are adjacent to the two lakes. The study did not attempt to
                                        produce independent estimates of such impacts but reported the results of
                                        research done elsewhere on related issues. Two of the cited studies
                                        suggest significant negative effects of the drawdown of lake levels on the
                                        value of adjacent properties.


                                        Both LOUD and TVA representatives told us that the study had several
LOUD and TVA                            weaknesses. LOUD representatives stressed that the benefits reported in the
Pointed to a Number                     study were “low” estimates of potential benefits. The authors of the study
of Weaknesses in the                    agreed, citing that due to limited funding, the study suffered from
                                        limitations that probably resulted in underestimating benefits. LOUD
Cherokee and                            representatives pointed out, for example, that one estimate of the impact
Douglas Study                           of delays in the drawdown of lake levels in the study is based entirely on
                                        the respondents to the commercial establishment survey. The response
                                        rate to the survey was quite low (about 18 percent), and the study assumed
                                        that only the respondents’ retail sales would increase due to delaying the
                                        drawdown of the lakes and that that nonrespondents would not
                                        experience increases in retail sales. The study’s authors agreed that this
                                        approach could have under-estimated retail sales impacts. On the other
                                        hand, the authors pointed out that respondents would also have an
                                        incentive to inflate their estimates of retail sales increases due to delays in




                                        Page 85                            GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
    Appendix IV
    Study Examining Cherokee and Douglas
    Lakes




    the drawdown of lake levels in the hopes of influencing a decision to
    maintain higher lake levels.

    LOUD representatives emphasized that the most important benefit to the
    local economy would not be the increased visitation to the lakes, but
    rather to an increase in the re-location of retirees and others from outside
    the region. The representatives believe that the economic implications of
    this migration to the region are likely to far exceed the estimates reported
    in the study.

    TVA had a number of concerns regarding the Cherokee and Douglas Study,
    including, but not limited to, the fact that

•   the study estimates the impacts of delayed drawdown only on a six-county
    region as opposed to impacts on the entire TVA region and the United
    States as a whole. However, TVA noted that delays in the drawdown of lake
    levels could affect flood control and navigation on the Tennessee, Ohio,
    and Mississippi Rivers. In addition, TVA must consider impacts on
    hydroelectric power generation, flood control, water quality, and water
    supply as they affect the nation, the broader TVA region, as well as the
    six-county region. For example, TVA’s 1990 review quantified potential
    impacts on hydroelectric power generation revenue and considered
    potential flood control and navigation impacts on the lower Ohio and
    Mississippi Rivers;
•   the results of the study’s survey of lake visitors suggest that the benefits
    claimed for the six-county region could be shifted from the immediate
    surrounding areas. In other words, benefits to the six-county region could
    be partly at the expense of neighboring counties;
•   the study does not deal with possible increases in costs associated with
    expanding the six-county region’s infrastructure to accommodate
    increased visitation and economic growth; and
•   increased employment may be in the form of temporary importation of
    laborers into the area, with little effect on local unemployment.

    TVAalso raised some methodological concerns about the study, such as the
    accuracy of some of the survey results and the rigor of the statistical
    analysis of the effect of lake levels on retail sales.




    Page 86                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                        Appendix IV
                        Study Examining Cherokee and Douglas
                        Lakes




                        The authors of the Cherokee and Douglas Study intended to evaluate only
The Cherokee and        economic benefits to the six-county region of proposed changes to
Douglas Study Is        policies impacting lake levels. In contrast, federal guidelines for economic
Limited to the          evaluation of federal water projects recommend a more comprehensive
                        evaluation of costs and benefits of actions related to water projects.5 Both
Evaluation of           TVA and the authors of the study have recognized these guidelines as
Economic Benefits of    relevant for evaluating the changes to policies impacting lake levels under
                        consideration.
the Six-County Region
                        According to the guidelines, an evaluation of plans related to water
                        resources requires the estimation of expected changes in the economic
                        value of national output of goods and services from a plan. Although the
                        guidelines recommend consideration of other costs and benefits, such as
                        environmental costs and benefits, its only required estimation is changes
                        in the economic value of national goods and services produced.

                        Expansion of the study’s analysis to areas beyond the six-county region in
                        accordance to these guidelines could result in substantially different
                        estimates of benefits. This difference would be especially true if the major
                        portion of the study’s estimated gains in the six-county region is due to
                        transfer of economic activities into these counties from outside the region.
                        At the national level, inter-regional transfers of economic activities, in
                        general, result in no net change in the value of output of goods and
                        services produced and result in no national benefit.

                        Consistent with its stated scope, the study does not attempt to
                        independently quantify any other costs and benefits resulting from the
                        proposed delay of the drawdown of lake levels.6 It should be noted,
                        however, that the study’s estimates of economic benefits pertain only to
                        the six-county region surrounding the lake. Because of their limited
                        geographical scope, such benefit estimates are not comparable to TVA’s
                        estimate of its systemwide cost of supplying electric power.




                        5
                         U. S. Water Resource Council, Economic and Environmental Principles and Guidelines for Water and
                        Related Land Resources Implementation Studies, (Mar. 10, 1983). These guidelines establish standards
                        and procedures for use by federal agencies in the formulation and evaluation of alternative plans for
                        water related studies.
                        6
                         The Cherokee and Douglas Study reports TVA’s 1990 estimate of the annual impact on TVA’s
                        systemwide cost of supplying electric power at $18.6 million (adjusted from the 1990 figure of
                        $15 million for inflation to reflect 1998 values).



                        Page 87                                GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix V

Study Examining Georgia Mountain Lakes


                        This appendix gives a brief description of the study entitled, The
                        Economic Impact of Alternate TVA Lake Management Policies (henceforth,
                        the Georgia Mountain Lakes Study), which was prepared for the Mountain
                        Lakes Study Committee by the North American Water Management
                        Institute, Inc. As mentioned in chapter 4, this study focused on estimating
                        economic benefits to areas bordering three TVA lakes—Blue Ridge,
                        Chatuge, and Nottely. The study also analyzed options for funding TVA’s
                        increased cost to its power program due to changes to lake level policies.
                        This appendix also summarizes views of the study expressed by TVA and
                        our comments on the study.


                        The Georgia Mountain Lakes Study estimated very favorable benefit-cost
The Georgia Mountain    ratios for delaying the drawdown of lake levels at Blue Ridge, Chatuge,
Lakes Study             and Nottely Lakes in northern Georgia from August 1 to October 1. In
Estimated Favorable     addition to benefit-cost estimates, the study focuses on options for funding
                        TVA’s increased cost to its power program due to changes to policies
Benefit-Cost Ratios     impacting lake levels.
for Delaying
                        Similar to the Douglas and Cherokee Study, the Georgia Mountain Lakes
Lake-Level Drawdown     Study did not estimate the impact on TVA’s systemwide cost of supplying
at Three TVA Lakes in   electric power resulting from the drawdown of lake levels. Instead, the
Georgia                 study applied simple modifications to cost estimates reported in TVA’s 1990
                        review, obtaining an estimated increase in TVA’s systemwide cost of
                        supplying electric power of $750,000. The study’s benefit estimates relied
                        on 1990 TVA and U.S. Forest Service (USFS) visitation statistics for the three
                        lakes and a 1990 USFS lake visitors’ expenditures survey for the Blue Ridge
                        and Hiwassee Lakes. The study extrapolated the visitation data to 1995,
                        and applying the expenditures data, computed “incremental impacts of
                        lake visitation” due to delaying the drawdown of the lakes. The benefits to
                        Fannin, Union, and Towns counties were reported as ranging from
                        $3.7 million to $24.3 million for delaying drawdown to October 1, and from
                        $4.6 million to $30.2 million for delaying drawdown to October 31. The
                        study combined its benefit and cost figures to compute highly favorable
                        benefit-cost ratios as indicated in table V.1.1




                        1
                         The benefit-cost ratio is a measure of the proportion of benefits to costs and is derived simply by
                        dividing the former by the latter. For example, a benefit-cost ratio of 40.0 suggests that benefits exceed
                        costs 40 times.



                        Page 88                                 GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                                           Appendix V
                                           Study Examining Georgia Mountain Lakes




Table V.1: Incremental Impacts of Lake Visitation for Three Georgia Mountain Lakes
                              Based on figures           Based on figures        Based on figures                Based on figures
                           extrapolated from 1990 extrapolated from 1990 extrapolated from 1990               extrapolated from 1990
                             TVA lake visitation        USFS lake visitation    TVA lake visitation            USFS lake visitation
  Economic measure                statistics                 statistics             statistics                       statistics
                             Delay until October 1      Delay until October 1      Delay until October 31      Delay until October 31
Benefits                                $3,681,646                 $24,278,817                  $4,611,715                 $30,176,781
Costs                                      $750,000                   $750,000                    $750,000                    $750,000
Benefit-cost ratio                               4.9                       32.4                         6.2                        40.2
                                           Source: The Economic Impact Of Alternate TVA Lake Management Policies, North American
                                           Water Management Institute, Inc., (Dec. 1997).



                                           The study also focused on different methods of funding the increased
                                           costs of TVA power supplies associated with this change to policies
                                           impacting lake levels. Ten funding mechanisms were identified and
                                           discussed. All 10 funding mechanisms are of a local nature, in the sense
                                           that those who pay are either county residents or users of services offered
                                           within the counties. These options include surcharges on county
                                           customers of electric supply companies, local boating fees and fishing
                                           licenses, hotel-motel taxes, and a special assessment tax for lakefront
                                           properties. The study computed various measures of impacts of the
                                           funding alternatives, such as increases in individual electric customers’
                                           monthly charges; increases in boat licenses and fees; increases in
                                           hotel-motel tax rates; and increases in lakefront property tax rates. The
                                           Georgia Mountain Lakes Study discussed the various options with respect
                                           to impacts on payees and political feasibility.


                                           TVAhad a number of concerns regarding the Georgia Mountain Lakes
TVA Criticized the                         Study. According to TVA:
Georgia Mountain
Lakes Study on                         •   The study did not present sufficient information on its data, assumptions,
                                           and methodology to permit a professional review.
Several Counts                         •   The study used “very old” data. TVA also suggested that data projections
                                           were done in a way that inflated benefits estimates.
                                       •   The study failed to address other benefits and costs adequately. The study
                                           focused only on benefits to the three Georgia counties and only on the
                                           impact on TVA’s systemwide cost of supplying electric power. For example,
                                           the study did not consider possible adverse effects of drawdown delays on




                                           Page 89                            GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
                            Appendix V
                            Study Examining Georgia Mountain Lakes




                            navigation and flood protection.2 The study failed to account for
                            infrastructure and land development costs that would have to be made to
                            accompany the projected increase in lake visitation. TVA also criticized the
                            study for failing to properly recognize the multipurpose role of TVA’s
                            system of projects and the interdependencies among the purposes of the
                            projects.
                        •   The study did not take into consideration that some of the increased
                            visitation and estimated benefits might be transfers from other activities
                            that visitors would otherwise have chosen.
                        •   In discussing options for local funding of TVA’s increased electricity
                            production costs, the study fails to recognize that some local residents
                            would not benefit from lake drawdown delays. According to TVA, the study
                            fails to address the ability of the local population to compensate TVA for
                            the increased electricity production costs.


                            The cost and benefits estimation methodology used in the Georgia
The Georgia Mountain        Mountain Lakes Study does not conform to recommended federal
Lakes Study Uses            guidelines for evaluation of major actions at federal water projects. The
Simplistic Estimates        study’s benefit estimates are simply measures of increased expenditures,
                            which are not acceptable measures of economic benefits under federal
of Costs and Benefits       guidelines. In addition, the cost estimates used in the study are based on
of Lake Drawdown            very simple extrapolations from TVA’s 1990 cost estimates.
Delays                      On the cost side, the study relied on TVA’s 1990 estimates of the
                            systemwide cost of supplying electric power of Alternative 1D. In the 1990
                            study, Alternative 1D called for delaying drawdown at Blue Ridge, Nottely,
                            and Chatuge, as well as Hiwassee (not included in this study) from
                            Memorial Day to October 1, and for delaying drawdown at six other
                            multipurpose tributary projects from Memorial Day to August 1. TVA
                            estimated that this alternative would increase its annual systemwide cost
                            of supplying electric power by $3 million in 1990 dollars. The study then
                            reduced this amount by $2 million because TVA’s 1990 estimate of delaying
                            drawdown from Memorial Day to August 1 (for all 10 tributary lakes) was
                            estimated to cost $2 million annually. The study’s authors then prorated
                            the remaining $1 million over the four lakes, assigning $250,000 to each.
                            Because the Hiwassee project was not considered in the study, the cost of
                            the drawdown delay for the three remaining lakes was reported as

                            2
                             It should be noted that, according to TVA’s 1990 cost impact analysis, there does not appear to be
                            significant navigation and flood control impacts for Alternative 1D. This alternative delays the
                            beginning of summer drawdown from Memorial Day to October 1st for Hiwassee, Blue Ridge, Nottely,
                            and Chatuge Lakes, and to August 1st for six other tributary lakes. Because the study evaluates impacts
                            of drawdown delays from an August 1st base, as opposed to a Memorial Day base, the navigation and
                            flood control implications, if any, should be even less significant.



                            Page 90                               GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix V
Study Examining Georgia Mountain Lakes




$750,000. This cost calculation is over-simplistic because it ignores
changed conditions between 1990 and 1997, when the study was
conducted. Furthermore, the $750,000 was not adjusted for inflation.

The study’s estimated benefits to the three counties also do not conform
with federal guidelines. They are estimates of increased expenditures in
the three counties due to increased visitations by lake users. Not all
expenditure increases, however, can legitimately be counted as net
economic benefits to the three counties. The study fails to translate the
increased expenditure to net economic benefits. Similar to the Cherokee
and Douglas Study, the Georgia Mountain Lakes Study falls short of being
a comprehensive study of the costs and benefits of lake-level alternatives.




Page 91                        GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Appendix VI

Major Contributors to This Report


                       Phil Amon
Resources,             Margie Armen
Community and          Philip Farah
Economic               John P. Hunt, Jr.
                       Mehrzad Nadji
Development Division




(141250)               Page 92             GAO/RCED-99-154 TVA’s Multipurpose Tributary Projects
Ordering Information

The first copy of each GAO report and testimony is free.
Additional copies are $2 each. Orders should be sent to the
following address, accompanied by a check or money order
made out to the Superintendent of Documents, when
necessary. VISA and MasterCard credit cards are accepted, also.
Orders for 100 or more copies to be mailed to a single address
are discounted 25 percent.

Orders by mail:

U.S. General Accounting Office
P.O. Box 37050
Washington, DC 20013

or visit:

Room 1100
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting Office
Washington, DC

Orders may also be placed by calling (202) 512-6000
or by using fax number (202) 512-6061, or TDD (202) 512-2537.

Each day, GAO issues a list of newly available reports and
testimony. To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512-6000 using a
touchtone phone. A recorded menu will provide information on
how to obtain these lists.

For information on how to access GAO reports on the INTERNET,
send an e-mail message with "info" in the body to:

info@www.gao.gov

or visit GAO’s World Wide Web Home Page at:

http://www.gao.gov




PRINTED ON    RECYCLED PAPER
United States                       Bulk Rate
General Accounting Office      Postage & Fees Paid
Washington, D.C. 20548-0001           GAO
                                 Permit No. G100
Official Business
Penalty for Private Use $300

Address Correction Requested