Uncertainty in the Regional Portfolio Model Michael Schilmoeller for the Northwest Power and Conservation Council Generation Resource Advisory Committee Thursday, January 22, 2009

2 Overview  Construction Costs  Electricity Price Uncertainty  Economic Retirement  Variable Capacity for Existing Units

3 Attributes of Cost Futures  Generation Resource Advisory Committee (GRAC) at their December 18, 2008 meeting suggested futures should reflect  Expectations about cost trends  Greater uncertainty associated with technology that is immature or carries more political baggage  Correlation among costs, due to common elements such as steel, switches, labor, and concrete  Distinctions due to unique elements, likely reliance on sources outside the United States and associated exchange rate uncertainty, and so forth

4 Cost Components More Consistent Source: Forecasting construction costs.xls, worksheet MJS 2 (See also range “CERA Commodity Inputs” on wks MJS 3

5 Expectations Source: Random Variable 02.xls, worksheet “Graphs of Futures”

6 Cost Matrices Source: 2008 Nov 17 CCAF Power Tabular Data _ MJS.xls, worksheet “Table 2”

7 Input-Driven Behavior based on CERA Scenarios

8 Influences in Plant Cost correlations stochastic GBM wmr plant cost escalators steel cost growth expections stochastic GBM wmr stochastic GBM wmr auxillary equipment major equipment I/O matrix … … plant costs current overnight costs stochastic uncertainty in growth correlations with other plant costs stochastic growth

9 Individual Futures Source: Random Variable 02.xls, worksheet “Graphs of Futures”

10 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

11 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

12 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

13 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

14 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

15 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

16 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

17 Cost Deciles Source: Random Variable 01.xls, worksheet “Statistics”

18 Overview  Construction Costs  Electricity Price Uncertainty  Economic Retirement  Variable Capacity for Existing Units

19 Current Electricity Price Forecast  Interim, created Monday 1/5/2009  PNW prices east of the Cascades  Average hydrogeneration conditions  Most recent (12/20/2008) frozen efficiency electricity requirements forecast  Most recent (11/28/2008) natural gas price forecast  To be updated:  CO2 penalty assumptions  RPS resource development assumptions  cost and availability of new resources

20 Current Electricity Price Forecast Source: chart “MONTHLY PRICES”, wkb “AO_6P_ _ABL_HD MONTHLY POWER PRICE FORECAST.XLS”

21 Current Electricity Price Forecast Source: chart “MONTHLY PRICES2”, wkb “AO_6P_ _ABL_HD MONTHLY POWER PRICE FORECAST.XLS”

22 Current Conditions (ICE) Source: chart “Chart 1”, wkb “ICE MONTHLY AVERAGE XLS”

23 Current Electricity Price Forecast On- and Off-Peak Average Source: wks “Electricity Price Futures”, wkb “Illustrations and graphs from L805 after extraction.xls”

24 Current Electricity Price Futures Source: wks “Electricity Price Futures”, wkb “Illustrations and graphs from L805 after extraction.xls”

25 Deciles for Electricity Price Source: wks “EP Deciles”, wkb “Illustrations and graphs from L805 after extraction.xls”

26 Close-Up on Deciles Source: wks “EP Deciles”, wkb “Illustrations and graphs from L805 after extraction.xls”

27 Overview  Construction Costs  Electricity Price Uncertainty  Economic Retirement  Variable Capacity for Existing Units

28 Retirement Decision  Prescribed number of evaluation and mothball periods for each plant  If decision criterion, using only forward- going FOM is negative, begin evaluation  Once evaluation period is over, begin mothball state (no generation)  Once mothball periods have passed, decommission and incur cancellation cost  If any period yields a positive decision criterion value, reset all the counters above

29 Issues  What kind of costs, in addition to normal operating costs, would be incurred in each phase?  Is there a better representation of the retirement decision?

30 Overview  Construction Costs  Electricity Price Uncertainty  Economic Retirement  Variable Capacity for Existing Units

31 New Capability  Only for existing plants with a single cohort  Variable capacity could be stochastic  Representing uncertainty about whether anticipated capacity will materialize  Each future’s capacity different  Can increase or decrease during over the study within a future  Capacity would affect only FOM costs  Could be used to model RPS development uncertainty, if restricted to the case of increasing capacity over the study

32 Issues  Are there other aspects of a plant with varying capacity, such as construction cost?

33 End

34 CERA Cost Components Source: 2008 Nov 17 CCAF Power Tabular Data _ MJS.xls, worksheet “Table 2”

35 CERA Cost Components Source: 2008 Nov 17 CCAF Power Tabular Data _ MJS.xls, worksheet “Table 2”

36 Existing Features  Market-Driven Planning  A feature we do not use for our planning: unplanned market additions, at a higher price  Early construction – response to circumstances  Existing decision based on market viability and prospective adequacy of the system  Mothball and cancellation options  Option to incur all expenses in the first period

37 New Features  Overnight construction cost uncertainty  Applies to mothball and cancellation costs as well  Distinct costs if mothballed or cancelled in the first period of early construction  Fixed operations and maintenance (FOM) cost uncertainty  Capacity uncertainty  Decision criterion re-evaluated with revised going-forward cost only

38 New Features  Economic retirements  Based on covering FOM, optionally FOM with uncertainty  Internal, cohort-specific calculation of decision criterion  VOM uncertainty optional  Uncertainty in availability for immature technologies, transmission completion, u.s.w.  Distinct construction cost rates in early and late phases  Cohort-specific stochastic forced outages