Quantitative Viability Analysis of Electricity Generation Robert - - PowerPoint PPT Presentation

Quantitative Viability Analysis

• f Electricity Generation

Robert Riebolge/David Lenton

Carisway/DGA Consulting TUESDAY, 6 OCTOBER 2015

Agenda

• Overview of Project Objectives and Approach
• Approach to Forecasting SA Demand and Generation Mix
• Economic Model
• High Level Principles
• Costs and Benefits Assessed
• Sensitivity and Monte Carlo Analysis

Project Objectives

• Quantify the economic viability of a nuclear generator being

commissioned in South Australia in 2030 or 2050.

• Compare the cost effectiveness of market entry using a

nuclear generator against other generation options

• Small Nuclear
• Large Nuclear
• CCGT with CCS
• CCGT
• Consider the viability against a range of scenarios of

demand and renewable generation

• Produce a flexible and transparent model that allows the

user to modify the assumptions and consider the impact on the relative NPVs

Overview of Approach

Assessment and projection of demand for South Australia Forecast of renewable generation in South Australia Calculation of

• utput for new

South Australian generators Generator Assumptions & Scenario Selection Determination of costs & benefits applying per generator option Monte Carlo analysis on NPV to assess sensitivity Integrated model so changing demand/renewable scenarios will flow through to NPV Calculations

Approach to Forecasting South Australian Demand and Energy Generation Mix

Overview of Approach

• Historic Data Sets
• Data sets of demand and renewables generation measured at

half hourly (HH) intervals courtesy of SA Power Networks were provided for:

Demand

• Major customer category
• Business consumer category
• Residential consumer category
• Hot water load

Renewables

• Photovoltaics (PV)
• Wind generation

Overview of Approach

Example of HH daily load data for the business consumer category in 2012/13

SETTLEMENT DAY SETTLEMENT PERIOD

Jul-12 Aug-12 Sep-12 Oct-12 Nov-12 Dec-12 Jan-13 Feb-13 Mar-13 Apr-13 May-13 Jun-13 1 0:00:00 507082 578654 524685 465071 571891 583575 499259 574170 572810 467666 570179 553491 1 0:30:00 496440 566402 505719 460541 559030 573235 492648 564146 563246 459399 553314 537788 1 1:00:00 489695 559181 494427 457628 549925 561510 481977 557341 556914 452459 548100 522420 1 1:30:00 480606 544134 484389 447330 546619 546904 473357 549322 551643 445206 534263 510946 1 2:00:00 475167 537159 478806 441106 540922 530539 464779 538243 543261 439085 528751 500751 1 2:30:00 462583 531623 470411 437980 535984 521377 460776 535650 540686 436649 527863 492770 1 3:00:00 454213 524770 466051 433732 536863 517812 453477 536936 540353 436977 518042 486969 1 3:30:00 455391 523095 464629 432411 540299 516999 452100 542572 540232 437005 514297 485230 1 4:00:00 452867 528992 467470 436312 550543 516919 455822 552739 548939 442678 517786 480005 1 4:30:00 450683 535909 469357 436534 572372 527896 456916 573281 571621 450799 525521 481412 1 5:00:00 452535 542214 468835 442049 607709 535297 456129 603187 603235 457231 534028 481833 1 5:30:00 459366 566121 480612 452311 644801 529268 437724 663376 664978 476638 562895 494358 1 6:00:00 464939 607098 497522 454144 683427 547203 434046 688732 718997 485854 603432 507606 1 6:30:00 479257 683664 529447 453882 745953 575117 448299 743211 746891 504232 665180 537413 1 7:00:00 487486 756787 519000 469009 821885 605200 466124 800873 799238 486732 716740 561926 1 7:30:00 490673 825129 534619 499321 880889 639228 489015 860887 860637 495157 765764 574440 1 8:00:00 472987 903425 553369 509804 933716 673315 506314 893016 898397 505357 832204 574836 1 8:30:00 481935 1005868 577027 521405 963737 701609 525823 923657 927000 515781 894295 593576 1 9:00:00 489626 1066498 598456 523369 962343 716294 542829 924838 936058 525005 931372 610557 1 9:30:00 512330 1092114 612558 529623 967458 726097 562693 932636 949795 530297 947160 630920 1 10:00:00 523878 1077315 617125 530836 975759 732064 576610 945169 963996 536516 954461 635699 1 10:30:00 538215 1058719 618677 534846 970829 728182 595846 945821 970833 546716 954952 642677 1 11:00:00 557204 1045126 613910 545086 969228 727151 606394 947388 981928 545390 959151 644361 1 11:30:00 562186 1034526 608176 550989 963698 711294 612470 944628 987891 544842 962967 643671 1 12:00:00 561248 1014847 598519 550192 963002 698086 620978 947845 997304 552430 969904 635944 1 12:30:00 553519 993504 587914 552106 964556 690603 623571 940958 1005854 557686 965432 629090 1 13:00:00 555154 983315 581786 547936 961584 686354 627230 939092 1019039 555388 966033 622762 1 13:30:00 553154 968006 576419 543274 956411 680173 623325 928193 1027273 551350 964206 620539 1 14:00:00 548560 956090 568683 537917 947085 678314 625351 915077 1023827 545764 967742 614579 1 14:30:00 548420 946522 563206 538362 935491 675289 625257 893826 1010923 539620 967033 605452 1 15:00:00 544874 933832 557846 536994 909947 670179 626391 872773 991477 537704 952232 599789 1 15:30:00 543945 911346 548609 533820 879498 657431 625738 845751 970353 537979 933185 592001 1 16:00:00 541662 889198 544836 530906 849541 647759 622058 818470 941397 533487 909367 591812

BUSINESS DEMAND (kW) 2012/13

Note: This data is sourced from SA Power Networks

Overview of Approach

• The importance of ‘Big Data’
• Data sets of demand and renewables generation measured at

half hourly (HH) intervals provide for:

• More accurate categorisation of load characteristics of:
• Peak load values and durations
• Temporal (i.e. week days or weekends) and seasonal variability
• Disaggregated consumer and renewable generation

characteristics

• Finer HH granularity meaning load shapes more closely mimic

real time load shapes

• Greater confidence in forecasting load shapes
• ‘Statistically’ more credible forecasts

Historic System Demand

Residential load curves at HH granularity in 2012/13

Historic System Demand

Business load curves at HH granularity in 2012/13

Historic Renewable Generation

PV generation at HH granularity in 2012/13

Historic Renewable Generation

Wind generation at HH granularity in 2012/13

Fossil Fuel Generation

Fossil fuels at HH granularity in 2012/13

Overview of Approach

• Historic Demand by Consumer Category
• The following figures display demand profiles for a minimum

demand day and a maximum demand day for:

• Major customer category
• Hot water load
• Business consumer category
• Residential consumer category
• Note that the business and residential consumer categories

make up the bulk of the system load in South Australia.

Historic System Demand

System daily load curves for a minimum and maximum demand day in 2012/13

Historic Demand by Consumer Category

Consumer category daily load curves for a minimum and maximum demand day in 2012/13

Overview of Approach

• Historic Renewables Generation
• The following figures display generation profiles of

renewables by season for:

• Photovoltaics (PV)
• Wind
• Note that PV generation is highly seasonal while Wind does

not have the same degree of variability by season.

Historic PV and Wind Generation by Season

PV and Wind generation for a week in summer and winter in 2012/13

Historic PV and Wind Generation by Season

PV and Wind generation for a month in summer and winter in 2012/13

Overview of Approach

• Historic Generation to Meet the System Demand
• The following figures display generation profiles for:
• Photovoltaics (PV)
• Wind
• Fossil Fuels

needed to meet the system demand for a minimum demand day and maximum demand day.

• Note that on low demand days much of the system demand is

met by renewables.

Historic Generation to Meet the System Demand

PV generation for a minimum and maximum demand day in 2012/13

Historic Generation to Meet the System Demand

Add wind generation for a minimum and maximum demand day in 2012/13

Historic Generation to Meet the System Demand

Add fossil fuel generation for a minimum and maximum demand day in 2012/13

Forecasting System Demand in 2030

• The model’s ‘Variables Input Sheet’ for demand

parameters to be varied provides for the following:

• Business consumer category (high, low, medium).
• Residential consumer category (high, low, medium).
• Major customer category (high, low, medium).
• Hot water load (high, low, medium).
• Co-generation (yes, no).
• Electric vehicles (% of vehicle population).

Example Inputs to the Demand Model in 2030

Business (% pa) medium 1.2% Residential (% pa) medium 1.5% Major Customers (% pa) medium 0.2% Hot Water Load (% pa) low

• 0.2%

Co-generation (on/off) no Electric Vehicle Market Share (%) 25% Growth in Demand

• The demand parameters are applied to the HH data sets for:
• Business consumer category
• Residential consumer category
• Major customer category
• Hot water load
• Co generation is set either ‘on’ or ‘off’
• A percentage of vehicle market share can be chosen for EVs

Scenario Forecast System Demand in 2030

System demand in the SA grid in 2030

Scenario Forecast System Demand in 2030

Major customer category daily load curves for a minimum and maximum demand day

Scenario Forecast System Demand in 2030

Adding hot water load daily load curves for a minimum and maximum demand day

Scenario Forecast System Demand in 2030

Adding business category daily load curves for a minimum and maximum demand day

Scenario Forecast System Demand in 2030

Adding residential category daily load curves for a minimum and maximum demand day

Scenario Forecast System Demand in 2030

Adding electric vehicle (EV) daily load curves for a minimum and maximum demand day

Forecasting Generation in 2030

• The model’s ‘Variables Input Sheet’ for generation

parameters to be varied provides for the following technologies:

• PV penetration of business consumers (high, low, medium).
• PV paired with storage (high, low, medium).
• Wind paired with storage (high, low, medium).
• Wind installed capacity (high, low, medium).
• Solar Thermal Plant installed capacity (high, medium, low).
• Nuclear Plant installed capacity (high, low).
• Interconnector constraint (high, medium, low).
• Vehicle to Grid (V2G) penetration (high, medium, low).

Example Inputs to the Generation Model in 2030

Penetration of Business Category medium 30% Penetration of Residential Category saturated PV's paired with Storage high 80% Wind paired with Storage high 60% Wind Installed Capacity MW) medium 2000 STP Installed Capacity (MW) medium 200 Interconnector Constraint (MW) medium 650 V2G Penetration medium 40% Solar Thermal Plant (STP) Interconnector Constraint Technology Forecast Photovoltaics (PV) Wind Generation Vehicle to Grid (V2G)

• The above technology parameters are applied to the historic HH data sets for

PV and wind generation and appropriately scaled up

• Installed capacities of STP, Nuclear and the Interconnector Constraint are

determined by the values chosen and these values set the maximum power

• utputs and constraints

Scenario Forecast of Generation Mix in SA in 2030

System demand in the SA grid in 2030

Scenario Forecast of Generation Mix in SA in 2030

Photovoltaics (PV) alone in the SA grid

Scenario Forecast of Generation Mix in SA in 2030

Dispatching photovoltaics (PV) paired with batteries in the SA grid

Scenario Forecast of Generation Mix in SA in 2030

Dispatching wind generation in the SA grid

Scenario Forecast of Generation Mix in SA in 2030

Dispatching wind generation paired with grid storage in the SA grid

Scenario Forecast of Generation Mix in SA in 2030

Dispatching V2G release from electric vehicle storage in the SA grid

Scenario Forecast of Generation Mix in SA in 2030

Dispatching Solar Thermal Plant (STP) generation in the SA grid

Scenario Forecast of Generation Mix in SA in 2030

Dispatching Nuclear in the SA grid

Scenario Forecast of Generation Mix in SA in 2030

Dispatching fossil fuels in the SA grid

Forecasting Generation Export in 2030

• Using the same scenario of demand and technology

inputs:

Business (% pa) medium 1.2% Residential (% pa) medium 1.5% Major Customers (% pa) medium 0.2% Hot Water Load (% pa) low

• 0.2%

Co-generation (on/off) no Electric Vehicle Market Share (%) 25% Growth in Demand Penetration of Business Category medium 30% Penetration of Residential Category saturated PV's paired with Storage high 80% Wind paired with Storage high 60% Wind Installed Capacity MW) medium 2000 STP Installed Capacity (MW) medium 200 Interconnector Constraint (MW) medium 650 V2G Penetration medium 40% Solar Thermal Plant (STP) Interconnector Constraint Technology Forecast Photovoltaics (PV) Wind Generation Vehicle to Grid (V2G)

Total Power Available for Export in 2030

There is no export of PV alone, PV paired with storage and wind alone

Total Power Available for Export in 2030

Export wind paired with storage with no interconnector constraint

Total Power Available for Export in 2030

Export of V2G release from electric vehicle storage with no interconnector constraint

Total Power Available for Export in 2030

Export of Solar Thermal Plant (STP) power with no interconnector constraint

Total Power Available for Export in 2030

Export of Nuclear Plant power with no interconnector constraint

Total Power Export in 2030

Export of Nuclear Plant power with interconnector constraint

Inputs to the Economic Model

• The Demand/Generation component of the model then

generates values of energy sent out to supply the SA grid and to export to the NEM as per the Table below.

SA NEM SA NEM CCGT with CCS (327MW) Small Nuclear (600MW) Large Nuclear (1200MW) CCGT (374MW) Energy Sent Out (GWh) in 2050 Last Dispatch Mode Third Dispatch Mode

Economic Model

Economic Model High Level Principles

• Economic Model focused on Commercial Viability of a nuclear

generator entering the market in either 2030 or 2050

• Does not consider any societal benefits/costs from the investment that are

not captured by the generator

• Assumes that the value of carbon is captured through some form of

carbon price

• Proxy for some mechanism that values carbon
• Impacts on the wholesale price and cost of some generation options
• Option of which Carbon Price track to apply
• User selection for inputs affecting generation requirements
• Renewable Generation
• Demand Scenario
• Dispatch mode options
• Plant Availability

Economic Costs and Benefits Assessed

• Major Cost Splits into the following areas:
• Capital Costs
• Variable and Fixed Operating and Maintenance Costs
• Fuel Costs
• Carbon Costs
• Decommissioning Costs
• Connection and Network Upgrade Costs
• Benefits Divided into two areas:
• Sales of Electricity into South Australian Market
• Sales of Electricity via the Interconnector

Details on each of these calculations are provided below

Capital Costs

• Built up from overnight $/kW capital cost for each option
• Includes cost for pre-construction
• Interest during construction calculations built up from profile of

expenditure

• Discount rates included separately for all generation options
• Model includes the option for the user to assess impacts of:
• Budget Overruns
• Delay in project completion

Operations and Maintenance Costs

• Fixed (FOM) costs included based on a $/kW net figure
• Almost all nuclear on-going costs classed as fixed
• Mixture of international and local costs
• Variable (VOM) costs included on a $/MWh sent out basis
• Parameter escalates in real terms over the course of the

model in line with EY/AETA estimates

Fuel Costs

• Nuclear fuel costs provided as $/MWh Sent Out
• Varies for Small/Large nuclear options
• Costs are in USD
• Gas price tracks provided as $/GJ
• Compared against estimates from recent public domain reports
• Significant movement in international gas prices
• Consumption of gas driven by plant efficiency
• Built in material learning curve improvements by 2030 and 2050

Carbon Costs

• CCGT based on calculation of tonnes of CO2 emitted and carbon

price

• Carbon price based on 450 PPM target
• Options in the model to apply different carbon price tracks
• CCGT with CCS includes two elements of cost
• Carbon sequestration for carbon that is captured
• Carbon cost for uncaptured % of emissions
• Assumption is that there are no carbon costs associated with

nuclear fuel

• Ignores carbon costs in the fuel mining and decommissioning cycle
• Consistent with treatment of other fuels
• Calculations also included for total carbon amelioration benefits

Connection and Network Costs

• Connection costs included for all plant
• Relatively small in total costs calculation
• Infrastructure Costs separately added for each option
• Transmission and Interconnector upgrade costs for Large

Nuclear Plant

• Consider a HV Transmission link from Victoria through South Australia
• Option to determine the level of contribution from the large

nuclear plant for the upgrade

Decommissioning Costs

• Cost for the nuclear plant assumed to apply at the end of the plant

life

• Escalation factor applied to the decommissioning cost
• Included costs estimate for dry storage based on $/MWh levy to

cover cost of facility

• Materiality based on the expected cost, life of the plant and

assumed discount rate

• Retirement costs have been included for gas plant for

completeness, but low materiality

Electricity Sales

• Sales split between SA and exported via the interconnector to the

NEM

• Adjustments made for Marginal Loss Factors
• Model can test different levels of interconnector capacity
• Differential price can be applied in line with market models results

Key Data Sources

Types of Parameter Main Sources Nuclear Costs and Operating Parameters WSP-PB CCGT and CCGT with CCS Costs and Operating Parameters CO2CRC, AETA, FGF, EY Discount Rates KPMG, EY Wholesale Prices EY, FGF Carbon Prices EY All model inputs will have a clearly defined sources Cross checking and the setting of ranges for key parameters will be done using a variety of recent public domain reports from AEMO, AETA, FGF , IEA, etc.

Sensitivity and Monte Carlo Analysis

Why do we Need Sensitivity Analysis?

• Uncertain values for key parameters within the model
• Partly reflects plants not commissioning until 2030/2050
• Key parameters are therefore set up with a most likely value and a

high/low value including

• Fuel cost variations
• Parameters relating to initial capital cost
• Efficiency of plant
• Variations in wholesale and carbon prices
• VOM and FOM costs
• Exchange Rates
• Range of likely values tested in the Tornado Diagrams and Monte

Carlo analysis to consider range of possible values

Example of Sensitivity Analysis

Illustrative example only awaiting updated inputs

• 10%

52.0% 10.0% 0.950

• 10%

1277.1 10% 5.5 50 7.0% 10% 57.0%

• 10.0%

0.975 10% 1044.9 0% 4.5 30 9.0%

Variation in Wholesale Price without Carbon (0%) Efficiency of CCGT in 2030 (55.1%) Percentage change in Gas Prices (0%) MLF Small Plant (0.975 ) % Change in Carbon Price from Most Likely Predictions (0%) Capital Cost of CCGT in 2030 ($1161/KW) Overrun of Budget for CCGT (0%) VOM CCGT ($5/MWh sent out) Life of CCGT (40 years) Discount Rate Real CCGT (8.78%)

NPV of CCGT in 2030 (M$AUD)

Showing Values >= 20.0 M$AUD

Monte Carlo – Example Output

NPV Example