Learning Center
Plans & pricing Sign in
Sign Out

TexMex Power Purchase Strategy


									                   The U.S. Power
                   2001 and Beyond
                   Dr. Arnold Leitner
                   RDI Consulting

August 13, 2001
The Golden Hotel
Current Drivers of US Electricity Markets
Understanding the drivers of the current US electricity market will tell you
where the market is headed

•   Power Plant Construction Boom
     - Expected Levels of Capacity Additions
     - Technology Mix
     - Retirements
•   Demand Growth
     - The Threat of a U.S. Recession
     - Power Prices
     - The Digital Economy
     - Interruptible Load
•   Demand and Supply Balance 2001-2003
•   Natural Gas – Panacea or Pandora
     - Price Forecast
•   Coal – The Black Work Horse
•   Gas-onCoal Competition
•   Renewables-on-Conventional Competition
•   The NEW Electricity Market

RDI Consulting                                                                 2
Construction Boom – The Numbers
Independent power producers and utilities add unprecedented amounts of new
generating capacity

                                          Historical and Forecast Capacity
   60,000                                             Additions

                                                     Utility               2001-2005 Owner %
   50,000                                            12%

   30,000                                                                                 Power
   20,000                                                                                  88%




































RDI Consulting                                                                                                        3
Construction Boom – The Picture
New power plants are being built everywhere in the nation

RDI Consulting                                              4
Electric Supply Outlook
There are 400,000 MW of proposed projects nationally
 Planned New Plant Additions To Meet Summer Peak Demand
                      Oper/Under Const.          Advanc. Dev.        Other    Total     Oper/Const.Adv
                   2001     2002    2003    2002      2003     2004                     as % of Peak Dmd.
 AZNMA/CANVA      3,672    3,966   5,139     727     1,978    3,455  33,605    55,084            29%
 ECAR/MAIN        7,145    6,913   1,697   2,287     3,472    1,836  73,583    97,458            18%
 ERCOT            6,311    5,441   3,030       77        26     550   9,379    24,814            28%
 FRCC             1,829    3,661   1,171     100      (582)   1,940  15,115    24,969            24%
 MAAC             1,440    1,210     830     -         -      1,020  15,876    20,980            10%
 MAPP               756      312     189     -         -        -     7,215     8,472            4%
 NEPOOL           1,680    4,752     720     540       272      -     2,381    10,345            34%
 NWPA             1,117      868   1,009       29      863      -    22,218    26,773            13%
 NYPP               328      123   1,080     110      (198)   2,022  12,359    16,074            11%
 RMPA               300      430     525     302       130      -     3,105     4,872            23%
 SERC             7,595   10,200   6,523     -       4,305    3,215  52,766    86,231            25%
 SPP              2,666    2,220     550     -         288        51 16,133    22,707            19%
 TOTAL           34,839   40,096 22,463    4,172 10,554 14,089 263,735        398,779

  •   7% of proposed projects are ―officially‖ cancelled.
  •   Since the construction boom began, more than 130,000 MW of capacity has begun operating or is
      under construction.
  •   ECAR/MAIN and SERC have the greatest amounts of early development capacity
  •   Of the 400,000 MW of proposed, how many projects will actually go forward?
       - Nationwide 26% of all proposed capacity is operating or under construction.
       - Highest project completion rates are around 45%.
       - In some regions only 15 to 20% of projects have moved forward thus far.
RDI Consulting                                                                                              5
    Expected Level of Capacity Additions
    Expected level of new capacity additions is 180,000 MW

                   Expected Level of New Capacity Additions (MW) To Meet Peak Dmd
                   Region          2001     2002     2003     2004     2005    TOTAL
                   AZNMA_CANVA     3,164    5,303    6,573    8,628    2,666    26,334
                   ECAR_MAIN       7,226    8,281    8,478   10,651    3,800    38,437
                   ERCOT           6,375    6,605    2,161      517      206    15,863
                   FRCC            1,910    3,796    1,304    4,466      873    12,349
                   MAAC              698    2,402    1,469    2,608    1,091     8,269
                   MAPP              788      583      429      116    1,977     3,893
                   NEPOOL          1,661    4,624    1,037      394       81     7,797
                   NWPA              720    1,536    1,466    2,485    1,032     7,239
                   NYPP              249      341    1,126    3,850      838     6,405
                   RMPA              295      686    1,014      538      348     2,880
                   SERC            6,983   11,219   10,343   10,474    2,466    41,485
                   SPP             2,487    2,206    1,653    2,207      569     9,123
                                  32,558   47,582   37,052   46,936   15,948   180,075

•     Of the total expected additions of 180,000 MW, 55% are already under construction or

•     The Midwest and Southeast account for 44% of total additions and 50% of the additions not
      yet under construction.

RDI Consulting                                                                                    6
More Capacity from Existing Power Plants
Capacity creep from the existing system is one source of new capacity

•   According to NERC, more than 1,000 MW of new capacity will be added to the grid through
    up-rates between March and August of this year,

•   RDI estimates that an additional 2,000 to 3,000 MW of capacity is available at existing nuclear

•   Coal plant uprates are less likely due to New Source Review (NRS) requirements
     - Uprates still occurring at plants that are installing scrubbers.
     - Uprates may be occurring without announcements.
     - New plant owners are focused on asset productivity.
     - Bush Administration has recommended that new source review standards would not apply to
       minor modifications at plants – could result in many more up-rates.

RDI Consulting                                                                                   7
    The Technology Mix of New Power Plants
    Expected level of new capacity additions is 180,000 MW

              Total 182,228 MW
                                   2%1%2%                                Coal
                                                                         Combined Cycle
                                                                         Combust Turbine

•    RDI Consulting forecasts 128,735 MW of Combined Cycle plants to come online in the near future.
     Many projects are still in early development.
•    Modern Combined Cycle plants use supplemental firing in the HRSG ducts to boost output by 10-
     30% during peak demand.
•    In high solar resource areas Parabolic Trough solar fields could provide steam augmentation,
     substituting for duct firing.
•    Tax credits, easier permitting of project, and fuel price hedge are advantages for the developer.
•    Potential market for Parabolic Trough, price to ―beat‖ is $200-250/kW.
    RDI Consulting                                                                                8
Electricity Demand Outlook
RDI’s forecast demand is slightly lower, but close to historic levels

•   Historical growth rate from 1995 to 2000.
•   Forecast growth rate from 2001 to 2005.
•   Forecast also includes one-time downward revision for 2001 in the West due to higher
    electricity prices.
Drivers of Demand – The Economy
Will a slowdown in economic growth affect demand growth?

•   Since the recession of 1991, annual growth in real GDP has averaged about

•   During the same timeframe, annual growth in total electricity demand has
    averaged about 2.7%.

•   The consensus among economists and other analysts is that real GDP growth
    will slowed by an average of between 0.5-1.0% through 2005.

•   How would an economic slowdown affect electricity demand growth?
     - In the short-term, an economic slowdown does not have a substantial impact on
       electricity demand, although impacts vary by region.
     - A prolonged economic slowdown can have a more substantial impact as structural
       shifts in the economy and population migration may occur.

RDI Consulting                                                                    10
Drivers of Demand – Retail Prices
Retail rate increases could have a more substantial impact on demand growth

•   Our analysis indicates that electricity demand is heavily influenced by price

•   History however may not provide as much insight as required to determine the
    impact of retail price changes on demand.
     - Volatile prices have increased the value some consumers may achieve by
       curtailing consumption to resell power.
     - Not since the 1970s have consumers experienced the rapid price increases of the
       magnitude experienced over the past year.

•   There is clear anecdotal evidence of significant consumption drops in California.
     - Key question is how much of this drop reflects long-term structural change and how
       much is due to ―crisis‖ mentality.
     - Not enough data available yet to know.

RDI Consulting                                                                       11
Drivers of Demand – The Digital Economy
Digital economy and electricity demand – The Myth

Mark Mills, in a study partially funded by a coal lobby group:

• ―Somewhere in America, a lump of coal is burned every time a book is
    ordered online…‖

• ―In 1998, 13% of total U.S. electricity consumed went to powering Internet-
    related activities.‖

• ―Within 10-20 years, the total electricity consumed by the Internet could
    account for as much as half of all power consumed annually in the U.S.‖

• ―While total electricity demand growth averages 2-3% each year, as much
    of half of this growth can be attributed to the increased power consumed
    by the Internet.‖

RDI Consulting                                                                 12
Drivers of Demand – The Digital Economy
Digital economy and electricity demand – The Reality

•   Researchers, including leading authorities at Lawrence Berkeley National Lab (LBNL),
    refute Mills’ claims

•   Statistics don’t support claims of rampant power demand growth attributable to Internet
    popularity growth.
     - DOE EIA: ―From 1985 to 1995, retail electricity sales grew at a rate of 2.6% per year…Since
        1995, the use of the Internet has increased dramatically, yet retail electricity sales have grown
        by 2.1% per year, 0.5% less than the previous 10 years…‖

•   Ordinary desktop PC average power consumption
     - Mills (et al.) estimate = 1,000 watts.
     - LBNL (et al.) estimate = 65-215 watts.
     - Other numerous technical disputes over assumptions.

•   A trend to lower power consumption
     - Penetration of low-power flat panel displays.
     - In 2000, laptop sales increased 21%, desktop 1.6%.
     - Chip cooling technology not keeping up with increasing heat load.
     - Limited storage capability of batteries in portable electronic devices.

RDI Consulting                                                                                        13
Drivers of Demand – High Density Loads
High-Density Electric Loads

•   HiDELs—also called Internet hotels, server farms, or data centers—run 24
    hours a day, 365 days a year, and require a substantially more reliable power
    supply than other grid users (up to 99.9999% reliability).
    This a Power Quality Issue.

•   These facilities draw loads much higher than homes, office buildings, and
    industrial facilities. Utilities are already having difficulty with their
    distribution systems serving these high-density loads.
    This a Distribution Issue.

•   From 1999 to 2000, the amount of space occupied by server farms rose from
    3 million to 9 million ft2. Analysts predict that by 2003, this number will rise
    to 25 million ft2 (equivalent to 125 ten-story office buildings).

    How will this affect the demand for megawatt-hours?

RDI Consulting                                                                         14
Drivers of Demand – High Density Loads
Overall impact on electricity demand, however, should not be substantial

 Total energy consumption of data centers
                                                                             • Assumption:
       1600                                                                  average data center
                                                                             computer room
       1400                                                                  power density is
                                                                             equal to 60 W/ft2.

                                                                             • Note: actual
                                                                             average data center

                                                                             computer room
        600                                                                  power density could
                                                                             range from 35 to 85
        400                                                                  W/ft2.


               1998         1999          2000         2001          2003

 Source: Jennifer Mitchell-Jackson, Master’s Thesis, UC-Berkeley, May 2001

RDI Consulting                                                                                 15
Drivers of Demand – Interruptible Load
Will price-responsive demand become the death of peak demand growth?

•   According to NERC, almost 4% of current demand is interruptible (almost 30,000
    MW of capacity).

•   At what price does this demand become interruptible?

•   How much of this demand is truly interruptible?
     - The regions with the greatest reliance on interruptible demand include SERC,
       Texas, Florida, and the Midwest (4 to 7% of total peak demand)

•   Among energy managers for large companies, survey results indicate that if
    energy managers were offered $1000 per MWh, as much as 18% of their load
    could be shed.

•   Based on our survey results, as much as 50 GW of demand could ultimately
    become curtailable at a price of $1,000 per MWh.

RDI Consulting                                                                        16
    Drivers of Demand – Conclusions
    Demand growth follows old trends, but may take new turns

    •   Overall GDP growth is still the driver of demand.
    •   Power prices have a big impact on consumption.
    •   Computers and the Internet do not spur demand growth.
    •   High Density Load pose challenges to power quality and distribution networks
        but not power supply.
    •   Interruptible (Dynamic) demand has the potential to result in significant
        reductions in peak demand through load shifting

             • For developer, demand growth forecast and load shapes are the some of the
               key factor that determine the size and technology of new capacity. This is true
               for renewables also.

•       Do not look at historical—or even worse—current peak demand prices when
        performing an economic analysis of a generating technology. Things are likely to

RDI Consulting                                                                                   17
Demand and Supply Balance
RDI Consulting’s Forecast

RDI Consulting              18
2001 Forecast Reserve Margin

RDI Consulting   Green Range above Target RM Red Below Target RM
2003 Forecast Reserve Margin

RDI Consulting   Green Range above Target RM Red Below Target RM
Supply and Demand Balance - Conclusions
The US power markets are likely to be characterized by 70,000 MW of excess

•   Demand growth is not outpacing supply additions. The industry has not had
    difficulties adding capacity. Only 1/3 of current construction activity is need in
    the future to keep up with demand growth.
•   By the end of the forecast period, nearly every region will have much more
    generating capacity than it needs. Nationwide we expect up to 70,000 MW of
    excess capacity.
•   Our forecast only calls for 180,000 MW to be built. This is in contrast to the
    400,000 MW proposed.

         • The window of opportunity to capture some of the market of needed
            capacity is not only short, it is most likely already shut.

•   Renewable projects will have to find niches in a market that is likely to be
    characterized by rock-bottom energy prices.

RDI Consulting                                                                       21
Overview Conventional Fuels
Price and availability of fuels are a key driver for future power supply

•   Natural Gas
     - Future prices forecast at $3-4 mm/Btu continue to make natural gas economic.
     - Negligible SO2 emissions.
     - Modern emissions control technology provides excellent NOx reduction.
     - Lowest CO2 emission of any fossil fuel in combined cycle combustion.

•   Coal
     - Most abundant of all fossil fuels.
     - Even at $3.5 mm/Btu natural gas prices, new coal plants are only competitive at low coal cost—mine mouth
     - Carbon emissions are big bootleg for coal addition.
     - Clean coal technologies are not cost competitive and contain large technology risk

•   Nuclear
     - Very high capital cost
     - Substantial development cost and risk. National and local opposition by citizens is the single most important
        reason for project failure in current environment.
     - Uranium cost is likely to climb.

•   Hydro                                                     • Understanding what drives technology
     - Large Hydro? Forget it!                                   and fuel choice is key to determining
     - Small Hydro? Maybe, maybe.
                                                                 what renewable technologies to pursue.

RDI Consulting                                                                                                    22
Coal – The Black Workhorse
What does the future of coal look like?

•   The potential exists for up to 20,000 MW of existing coal-fired capacity to be retired by 2010.

•   Most retirements will be smaller, older plants in the eastern U.S., where retrofitting new emission control
    equipment is likely to be prohibitively expensive.

•   New Source Review (NSR) could play a role.

•   Developers may view plant retirements as opportunities to build new coal plants with the latest
    technology, producing winners and losers among coal suppliers.

•   Offset examples
     - Wisconsin Energy plans to build a new coal plant at its Oak Creek site, while re-powering its older Port
         Washington plant with natural gas
     - A proposed lignite-fired power plant in MAPP would include the construction of a wind-powered facility

•   Developers are looking to take advantage of a window of opportunity to build new coal-fired generating
    capacity. The duration of this window could be short

RDI Consulting                                                                                                    23
    Natural Gas - Panacea or Pandora
    Natural gas is the fuel of choice, but forward prices are higher than in the past

•     The Promise of Natural Gas
       - Clean, lowest CO2 emissions per MWh.
       - Combined cycle and combustion turbines have low capital cost, are modular, and are fast to built.
       - Competitive fuel price $/mmBtu.

•     The year 2000 – A Promise Broken
       - May be remembered as the most volatile year ever for natural gas markets.
       - Gas prices to most end-users doubled or tripled.
       - Many western locations saw physical spot prices at $20/mmBtu for sustained periods and peak prices in
          excess of $61/mmBtu.
       - The clear role that natural gas was going to play as the ―Fuel of the Future‖ is now mired in uncertainty.
       - Industry claims of continuous growth in gas supply at moderate to declining prices now rings hollow in the face
          of tremendous volatility.
       - 2000 brought on in large measure by the convergence of gas and electric market.

•     Natural Gas Tomorrow
       - High prices have spurred drilling. But can supply keep up with growth?
       - Pipeline Contracting and Siting: Expansions are planned but will they be certified, approved and built?
       - The year-round demand of natural gas fired combined cycle plants changes the dynamics of the gas market.

    RDI Consulting                                                                                                     24
Natural Gas – Drilling
Natural gas supplies will increase with increased drilling, but enough?

                                                                       U.S. WELL DRILLLIN G
                                                         3,0 00

                                                         2,5 00
            Month ly Wel l Com plet io ns in t he U.S.

                                                         2,0 00

                                                         1,5 00                                                                               N ATURAL G AS

                                                         1,0 00

                                                                             O IL

                                                           50 0


                                                                  Jan- 9 8

                                                                                        Jul- 9 8

                                                                                                              Jan- 9 9

                                                                                                                                                         Jan- 0 0
                                                                                                                         Apr- 99

                                                                                                                                   Jul- 9 9

                                                                                                                                                                    Apr- 00

                                                                                                                                                                              Jul- 0 0
                                                                              Apr- 98

                                                                                                                                              O ct- 99

                                                                                                                                                                                         O ct- 00
                                                                                                   O ct- 98

RDI Consulting                                                                                                                                                                                      25
    Natural Gas – Short Term Forecast
    Natural gas is the fuel of choice, but forward prices are higher than in the past

•     Tight gas supplies and growing demand should support pricing in the $3.00 to $5.00/mmBtu range over the
      next three years.

•     Overall gas price levels are primarily responding to supply-based factors, while transmission congestion
      can produce regional price spikes.

                                                                                                                                                    H EN RY_H UB
                                         $9.00                                                                                                      N YC_H UB
                                         $8.00                                                                                                      AECO-C_AB
                                                                                                                                                    Linea r (H EN RY_H UB)
                 REAL 20 00 $ / MMBTU





















    RDI Consulting                                                                                                                                                                               26
    Natural Gas – Long Term Forecast
    Natural gas is the fuel of choice, but forward prices are higher than in the past
•     In our Base Scenario, total annual gas demand is forecast to reach 29 trillion cubic feet (TCF) by 2010.
        - This results in prices in the $3.00 to range after the gas market comes back into balance after 2003;
        - If demand follows a lower trajectory ( i.e.,27 Tcf in 2010 ), prices could be 20% lower. Could cause demand
           destruction in the industrial sector.
•     Long-term prices are not expected to stay at the current levels.
        - Prices above $4.00/mmBtu erode demand and make a variety of marginal gas sources like coal-gas and new
           LNG projects viable.

                            Henry    N ew York CHICAGO     Suma s,    AECO- C, TOPOCK,
               YEAR          Hub     City HUB    HUB         WA         AB       AZ         OPAL
               1998         $2.00      $2.35    $2.12       $1.27      $1.21    $2.13       $1.78
               1999         $2.15      $2.52    $2.22       $1.65      $1.71    $2.31       $1.99
               2000         $4.07      $5.87    $4.15       $4.32      $3.40    $4.29       $3.68
               2001         $5.62      $6.32    $5.73       $5.41      $5.23    $5.98       $5.32
               2002         $3.95      $4.49    $4.04       $3.71      $3.58    $4.15       $3.57
               2003         $3.32      $3.79    $3.42       $3.09      $2.98    $3.46       $2.97
               2004         $3.18      $3.61    $3.27       $2.93      $2.83    $3.30       $2.85
               2005         $3.17      $3.62    $3.27       $2.89      $2.79    $3.28       $2.79
               2006         $3.07      $3.54    $3.17       $2.79      $2.70    $3.15       $2.67
               2007         $3.09      $3.57    $3.20       $2.77      $2.68    $3.18       $2.72
               2008         $3.08      $3.54    $3.19       $2.75      $2.66    $3.17       $2.70
               2009         $3.16      $3.65    $3.29       $2.80      $2.72    $3.27       $2.77
               2010         $3.22      $3.73    $3.35       $2.84      $2.78    $3.33       $2.81

RDI Consulting                                                                                                     27
           Coal-on-Gas Competition
           Delivered price projections for Henry Hub and into the Northwest Power Pool

          Delivered Gas Price Forecast ($2000/mmBtu)                                            •   We forecast delivered gas prices in
                                                                                                    NWP to decline in real terms over
                                                                                                    the next two years, then level off at
          6.00                                                                                      around $3.00 per mmBtu through the
                                                                                                    end of the decade.
                                                                                                •   Coal projects appear to be attractive
          4.00                                                                                      if coal costs remain below $1.20 and

                                                                                                    gas prices are above $3.50 per
          3.00                                                                                      mmBtu.

          2.00                                                                                  •   However, at $3.00 per mmBtu gas
                                                                                                    prices and coal prices between $1.00
          1.00                                                                                      and $1.20 per mmBtu, a new coal
                                                                                                    plant competes with gas CC
                                                                                                    technology only at very high
                                                                                                    capacity factors.










                                                                                                •   Coal becomes increasingly
                                              Henry Hub       NWP                                   competitive as delivered prices
                                                                                                    decline from $1.00 per mmBtu.

          RDI Consulting                                                                                                               28
Renewables-on-Conventional Competition
Competitiveness means more than low cost

          • Renewable energy research always seems to look at lowering cost. However, the true goal
             should be to maximize profits.
             Example: 10% lower fuel cost at Eastern power plants resulted in lower production cost, higher
             dispatch, AND LOWER PROFITS!

•   Know your ―proven reserves‖ of wind, solar, geothermal, and biomass.
         – Don’t estimate, but calculate your resources using latest GIS technology.
         – State your results in numbers that the industry understands, e.g. summer and winter rating,
            availability, forced ―outages‖. Develop new and intuitive terms if necessary.

•   Intermittency is o.k. , but being unpredictable is not. How well can solar and wind be forecast? Analyze
    historic data and develop methodologies. Then prove that you can do it! Example: Wind forecast in

•   Intermittency as power supply problem is mitigated if generation is disperse. What is the statistical
    availability of distributed wind or solar farms? Analyze and then prove concept with actual projects.

•   Do not focus too much on performance. Less may be more. Example: H-class combustion turbines.

RDI Consulting                                                                                                 29
Renewables-on-Conventional (Continued)
Competitiveness means more than low cost

         • Short-term energy storage increases value of renewable resource in energy
                       – Heat storage for thermal solar
                       – Compressed air storage for wind?
    But intermediate or long-term storage may not be profitable. A pile of coal is long-term
    energy storage. Is this really your business?

•   Always remember that building conventional generating capacities is not a no-brainer
    either. Advanced combustion turbines have plenty of problems. Understanding where
    you are in relation to other generating technologies helps to understand your strengths
    and weaknesses.

•   88% of all new capacity is built by independent power producers that are—for the most
    part—exposed to market risk, need the public acceptance, and live and die with their
    investment choices. Find out what these IPPs are looking for in a technology.

•   Commodities are sold on price—and attributes. Electricity from green energy can be sold
    at a premium. Know that premium.

RDI Consulting                                                                                 30
Renewables-on-Conventional (Continued)
Competitiveness means more than low cost

         • Every new technology needs an incubation period. Analyze the factors that
            resulted in the success of wind power in Europe and the dramatic cost reductions
            that accompanied it.

•   Be realistic. Credibility is one of the biggest areas of improvement for renewable
    generating technologies. No one expects miracles and neither does anyone believe them.

•   The Kyoto Protocol is real. Planting trees won’t work and carbon sequestration may be as
    cost effective and realistic as shooting radioactive waste from nuclear power plants to the
    moon. Renewable energies have a competitive advantage. This is a window of
    opportunity that is open and widening every day.

RDI Consulting                                                                                31
The NEW Electricity Market
Renewable generating technologies need to fit into the market

•   Energy trading is an emerging market
         • There is no need to come up with ways to ―firm-up‖ intermittent renewables
           with technical solutions when replacement power is available for less money
           from a trader.
         • Buying replacement power from a 7,000 mmBtu/kWh combined cycle plant
           may not only be cheaper, but is probably more environmental, too.

•   Accounting for the cost of ancillary services
         • Improve your output forecast abilities. Work on ancillary service up to cost
         • Find some one who will provide ancillary services for you, because ancillary
           services is not your business. Your business is zero marginal cost.

•   Who pays for transmission services?
         • Renewable energies need transmission access, but so does coal.
         • Other states may follow the ERCOT system operator, which hooks up
            everyone and makes all members pay.

RDI Consulting                                                                            32
The NEW Electricity Market
Renewable generating technologies need to fit into the market

•   Market structures can have significant impacts on economics of
    intermittent resources

         • Need to understand the market structures of system operators, such as NY
           ISO, Cal ISO, and ERCOT.
         • What are the bidding rules and penalties for failure to deliver or over-delivery.
         • Must participate in setting up market structures to avoid to be shut out.

•   Project Financing

         • Nothing guarantees project financing more than a PPA with a credit worthy
         • Need to develop financial tools that hedge intermittent generators from
           unsteady cash streams so that lenders are comfortable. This may be
           cheaper than buying replacement power of employing technical solutions.
         • Technology risk is a problem for project financing. Nothing assuages fears
           of banks more than demonstration projects under private ownership that
           play in the market. Always seek industry partnerships.

RDI Consulting                                                                                 33
Questions and Answers

RDI Consulting          34
Power Plant Construction Boom
Imperfect information is a key driver of current trend in power supply business
    2000 Capacity Additions                                            Capacity Margin vs. Reserve Margin
                                     Nerc Implied Capacity Additions
         10,000                                                                                     Capacity Reserve
                                     Actual Capacity Additions
                                                                              Demand Supply Reserve Margin Margin

          6,000                                                        2001    56.4   67.1     10.7     15.9   19.0

                                                                       2002    57.9   73.8     15.9     21.5   27.5
                                                                       2003    59.5   75.5     16.0     21.2   26.9
                  ECAR/MAIN   SERC           SPP
                                                                       2004    61.1   78.3     17.2     22.0   28.2

•        Lack of historical information from which to draw inferences regarding trends today
           - Demand elasticities
           - Actual cost/value of interruptible demand
           - Market driven ―reserve equilibrium‖ levels
•        Great uncertainty and often unreliable information associated with new capacity additions
•        Improper use of information – example of capacity margin calculations

RDI Consulting                                                                                                  35
Ordered Turbines Will Need Homes
Other factors could contribute to reaching total supply estimate

   U.S. Turbine Orders (MW)                                    •   Between 1998 and 2000,
                                                                   approximately 190,000
   100,000                                                         MW of turbines were
                                                                   ordered for delivery to
                       Gas Turbines    Steam Turbines
    80,000                                                         the U.S.

    60,000                                                     •   Siemens-Westinghouse
                                                                   estimates that
                                                                   worldwide turbine
    40,000                                                         manufacturing capacity
                                                                   is approximately 80 GW
    20,000                                                         per year.

          0                                                    •   Nearly 320 GW of
                                                                   turbines could be










                                                                   delivered between 2000










    Source: McCoy Power Reports                                    and 2004 worldwide.

RDI Consulting                                                                          36
Outlook for Plant Retirements
Excess capacity will lead to increased retirements

•   RDI believes most retirements will be concentrated in Texas, the
    Northeast, and the West
     - 6,000 to 8,000 MW of retirements in Texas.
     - 1,000 to 2,200 MW of retirements in NEPOOL.
     - 1,000 to 5,000 MW of retirements in the West.

•   Key factors impacting the level of retirements
     - Regulations and cost of service ratemaking.
     - Potential fixed cost reductions.
     - Fragmentation of the market and willingness of one player to make the first
     - Perceptions of market recovery.
     - Environmental regulations.
     - Future gas prices.
     - Financial impacts of closing a plant.

RDI Consulting                                                                       37

To top