1

Policy Analysis & Development of Integrated Middle East

Regional Energy Markets

Brian H. Bowen, F.T. SparrowZuwei Yu, Muhammad Al-Salamah

Purdue University

8th Power Generation ConferenceDubai, October 6 to 9, 2002

2

Benefits from Power Pooling“Tight vs Loose”

• Economies of Scale• Lower Reserve Requirement• Load Diversity• Economies of Operation• Cooperative Stimulus

3

• Purdue’s SUFG, State Utility Forecasting Groupwas formed in 1985.

• Reason for SUFG: Marble Hill Nuclear Station• SAPP legal entity 1995, Purdue 1997 – 2001• Purdue starts with WAPP in 1999• WAPP legal entity in 2000• SAPP comes under SADC, Southern African

Development Community• WAPP is part of the ECOWAS, Economic

Community of West African States• PPDG, Power Pool Development Group, is

the international component of SUFG

Purdue International ~ SAPP & WAPP

4

Consider 4 Regions

• Mid-West USA• Southern African Power Pool, SAPP• West African Power Pool, WAPP• Middle East Power Pool, MEPP

& Role of Purdue University’s SUFG,State Utility Forecasting Group

5

Electricity Market Modeling& Cost-Price-Demand Feedback Loop

Initial Prices& MarketStructure

Growing CustomerDemand

Utility SupplyThermal/HydroTechnologies

RegulatoryFrameworkRatesFinance

EquilibriumPrices

InvestmentPlans

Demand/TransmissionReliability

Costs/Trade

Pricing Policy/Contracts/Spot Markets

QuantifyBenefits

Purdue’sSUFG

Models

6

Mid-WestUSA Model

Detailed Modeling

IndianaIllinoisKentuckyMichiganOhioWisconsin

7

Model Characteristics & Applications

• Cost Minimizing• Engineering & Economics• Generation & Transmission – Existing & Proposed• Trading of Energy & Power Reserves• Policy Priorities – Interdependence, Reliability,

Sharing Gains, Pricing• Long-Term Planning – 10 to 20 years• Costs of operation, fuels, investments• Provision of policy decision support tools • Large regional model, user friendly interface,

easy technology transfer

8

Short Run, Energy Trade Only

( ) ( ) ( ) ( )t i z

min c i,z PG i,z,t UEcost UE z,t UMcostUM z+ +∑∑∑

s.t.

(1)

(2)

(3)

(4)

(5)

( ) ( ) ( )( ) ( ) ( ) ( )i zp zp

PG i,z,t PF zp,z 1-PFloss zp,z UE z,t D z,t PF z,zp+ + = +∑ ∑ ∑

( ) ( )PG i,z,t PGinit i,z≤

( ) ( )PF z,zp PFinit z,zp≤

( ) ( ) ( )i

PGinit i,z A z D z,peak≥∑

( )( )

( ) ( )i

PGinit i,zUM z D z,peak

1 res i,z+ ≥

+∑

9

Where:A) Variables: PG(i,z,t) = power generation at i in z during t (MW)

UE(z,t) = unmet energy demand in z during t (MW)

UM(z) = unmet reserve requirement in z (MW)

PF(zp,z) = power flow from zp to z (MW)

B) Parameters: = cost/MW of generation at i in z ($)

= cost/MW of unmet demand/reserves ($)

= line loss from zp to z (%)

= demand in z during t (MW)

= initial capacities (MW)

= reserve requirement for i in z (%)

= peak demand in z (MW)

A(z) = autonomy factor for z (%)

( )c i,zUEcost,UMcost

( )PFloss zp,z

( )D z,t

( ) ( )PGinit i,z ,PFinit zp,z

( )res i,z( )D z,peak

10

Short-Run, Power and Reserves Traded

s.t.

(1)

(2)

(3)

(4)

(5)

( ) ( ) ( )( ) ( ) ( ) ( )i zp zp

PG i,z,t PF zp,z 1-PFloss zp,z UE z,t D z,t PF z,zp+ + = +∑ ∑ ∑

( ) ( )PF z,zp PFinit z,zp≤

( )( )

( )( )

( ) ( ) ( )i zp

PGinit i,z Fmax zp,zUM z D z,peak Fmax z,zp

1 res i,z 1 res zp,z zp

+ + ≥ ++ +

∑ ∑ ∑

( ) ( ) ( )i

PGinit i,z A z D z,peak'≥ ≥∑

Where Fmax(zp,z) = reserves held by zp for z.

( ) ( ) ( ) ( )i z t

min c i,z PG i,z,t UEcost UE z,t UMcost UM z+ +∑∑∑

( ) ( )PG i,z,t PGinit i,z≤

11

Long Run Model

( ) ( ) ( ) ( )

( )( ) ( )

( )

Yi z t

yy=1

Y Y

y 1 y

c i,z PG i,z,t,y UEcost UE z,t,y UMcost UM z,ymin

1+disc

crf expcost i,z PGexp i,z,y

1+discτ

τ= =

+ ++

∑∑∑∑

∑∑

Subject to:

(1) & (3) With y added in variables

(2)

(4)

(5)

( ) ( ) ( )y

=1

PG i,z,t,y PGinit i,z PGexp i,z,ττ

≤ −∑( ) ( )

( )( ) ( ) ( ) ( )

y

1

i zp zp

PGinit i,z PGexp i,z,Fmax zp,z,y UM z,y D z,peak,y Fmax z,zp,y

1+res i,zτ

τ=

++ + ≥ +

∑∑ ∑ ∑

( ) ( ) ( ) ( )y

i 1PGinit i,z PGexp i,z, A z D z,peak,y

τ

τ=

+ ≥∑ ∑

12

Where:New variables: PGexp(i,z,y) = MW added in y at i in z

New parameters:

= cost/MW of expansion at i in z

= discount rate for present value purposes

= capital recovery factor

( )expcost i,z

disc

crf

13

SAPP Background & Modeling•12 National Utilities within Power Pool• Long-Term Capacity Expansion Planning • Displacement of thermal generation with hydropower

SAPP Demonstration Total Costs(for 16 years is $11.474 Billion)

Total Optimal Variable Costs for the HorizonFuel 5.604O&M 0.923Water 0.419

Total $6.947bnTotal Optimal Expansion Costs for the Horizon

Thermal 2.651Hydro 1.187Transmission 0.690Unserved Energy 0.000

Total $4.527bnDemonstration results & not used for project evaluation

14

Capacity Expansion Period 2001-2004: (a) Major (4490 MW) increase in transmission capacity of

DRC/Taz/Zam/Zim/Bot/NSA “spine,” connect DRC/Ang; (b) Build DRC Inga 3 (3500 MW), NSA Camden (950 MW), CT in Taz Period 2005-2008: (a) Minor (846 MW) increase in transmission capacity of spine; (b) Recommission NSA Grootvlei (1140 MW), Camden (570 MW), CT

(750 MW), build SSA pumped storage (3000 MW), Nam Epupa (360 MW), CT in Taz

Period 2009-2012: (a) Expand Zim/Moz transmission link; (b) Build hydro in Zam (940 MW), Zim (500 MW), NMoz (1240 MW),

Taz (268.5 MW), Mwi/Ang (90 MW), SSA (P$ 1000 MW); build SC in NSA (890 MW), CT in NSA (250 MW), NUC (1000 MW) in SSA

Period 2013-2016: (a) Minor (300 MW) increase in spine transmission capacity; (b) Build hydro in Ang (280 MW), Zam (240), Taz (220), Mwi (90), Zim

(200); build SC in Zim (1800), LC in NSA (3954), CC in Nam (750)

SAPP Least Cost Expansion Plan

15

SAPP ExpansionResultsof 20-Year Planning Horizon, Country Autonomy Constraint Relaxed (all units in MW)

2,000

3,726

500300

2,000

2,750

2826

43

972

= new generation capacity

= new transmission capacity

NH = new hydroSC = small coalPH = pumped hydroGC = combined cycleGT = gas turbineLC = large coal

CC-152GT-143

NH-557

NH-120

CC-1046NH-262

NH-360CC-485

LC-1372PH-3000CC-6277

NH-3500

GT-1320NH-572SC-1800GT-840

GT-235CC-3552

LC-5615SC-1590GT-1222

GT-343

CC-406

138

16

Autonomy Factors (MW and MWh) for SAPP, 2000 to 2020

Two trade commodities ~ MWh (energy)& MW (reserves)

* All 12 countries in SAPP use same level of interdependence, ie 0% & 70%

* Values from model “September28.gms”

Free Trade, MW (AFP = 0%)

Limited Trade MW (AFP = 70%)

Free Trade, MWh (AFE = 0%)

$13.10* billion $13.66* billion

Limited Trade, MWh (AFE = 70%)

$14.67* billion $14.81* billion

17

WAPP Background & Modeling

• 14 National Utilities within Power Pool• Zone A & Zone B parallel development • Prioritize new international transmission projects• Role of West African Gas Pipe Line, WAGPL• Infrastructure development• Policy analysis training• ECOWAS data collection, training & management

18

WAPP Gains from Trade& Improved Reliability

Thermal & HydroReserve Margins

WAPPTotal CostFree Trade($ million)

WAPPTotal Cost

& Independence($ million)

PercentageCost Savings

WithFree Trade

0% 7904 9288 14.9%5% 7924 9394 15.6%10% 7949 9524 16.5%20% 8060 9990 19.3%

Difference0% - 20%

2% 7.6%

19

27

10

150289

9322

2697

12

20

200

224 785 985

15

1175

WAPP Transmission Expansions, 2002-2020ECOWAS Country Notation:Ben – Benin BFa – Burkina Faso ICo – Cote D’IvoireGam – Gambia Gha – Ghana Gui – GuineaGbi – Guinea Bissau Lib – Liberia Mal – MaliNga – Nigeria Ngr – Niger Sen – SenegalSLe – Sierra Leone Tog - Togo

20

Applications to a Middle EastPower Pool ~ MEPP

• What are MEPP objectives?• Integrated or go-it-alone generation initiatives • Role of initial GCC regional sub-pool?• Other regional sub-pools• Immediate infrastructure & policy issues?

Modeling will quantify the gains fromtrade within a regional MEPP and demonstrate

the benefits of greater integrative planning

21

Middle EastInterconnections

Central Asia

22

ESCWA~13 Countries #1. Bahrain* 2. Egypt 3. Iraq 4. Jordan 5. Kuwait*6. Lebanon7. Oman*8. Palestine9. Qatar*10. Saudi Arabia*11. Syria12. United Arab Emirates*13. Yemen

ECO~10 Countries14. Afghanistan15. Azerbaijan16. Iran 17. Kazakhstan18. Kyrgzstan19. Pakistan20. Tajikstan21. Turkey22. Turkmenistan23. Uzbekistan

# = Also AFESD Countries* = GCC~6 Countries

Geographic Scope?

23

21

17

1623

19102

12

1115

3

5 20

7

18

13

9

1

6 84 22

14

Middle EastMarket Topology

Key:Large circle > 10,000 MW; Medium circle = 4,000 to 10,000 MW:Small circle < 4,000 MW

How big? Central Asia

24

MEPP Market Integration Policies1. Umbrella Structures: GCC, ESCWA, AFESD,

ECO, other sub-pools?2. Generation & Transmission & how to

collect existing data?3. Plans for proposed new generation?4. Plans for proposed new transmission &

gas pipelines?5. Policy priorities, interdependency, free trade,

reliability, least costs?6. Restructuring, IPPs, Regulators, Pricing

25

Modeling MEPP Plans & So What Next?(a) Regional Policy Modeling(b) Coordinating Institutions(c) MEPP Infrastructure Development(d) Prioritize Policy Initiatives Analysis(e) Purdue & Regional Collaboration(f) Organization for Data Collection(g) Outline for Modeling Proposal

26

MEPP Modeling Proposal ~ After Dubai October 2002

To Whom? ~ Utilities, Governments,Agencies?

What to do? ~ Quantify Trade Benefits,Pool Plan, Assess Structures?

Where to meet/plan/train ~ Locations?When to start? ~ 2002/2003?