MDESRAP: a model for understanding the dynamics of ... - CiteSeerX

Int. J. Global Energy Issues, Vol. 23, No. 1, 2005

1

MDESRAP: a model for understanding the dynamics of electricity supply, resources and pollution Hassan Qudrat-Ullah School of Administrative Studies, York University, 4700 Keele Street Toronto, Ont. M3J 1P3, Canada E-mail: [email protected] Abstract: The inherent dynamic complexity of the energy policy design problem makes the conventional ‘closed-form’ solution impossible. This paper contributes a dynamic simulation model as a possible solution. The model captures the dynamics of underlying sectors of electricity demand, investment, capital, resource, production, environment and costs and pricing. The existing feedback loops, nonlinearity, and time-lag characteristics present in the real world electricity systems are incorporated in the model. The model is calibrated to Pakistan’s case data. How the model has been used in policy assessments and the design of the alternative energy policies subject to various environmental and resource constraints is also discussed. Keywords: CO2 emissions; dynamic simulation model; Pakistan; system dynamics. Reference to this paper should be made as follows: Qudrat-Ullah, H. (2005) ‘MDESRAP: a model for understanding the dynamics of electricity supply, resources and pollution’, Int. J. Global Energy Issues, Vol. 23, No. 1, pp.1–14. Biographical notes: Dr. Hassan Qudrat-Ullah is an Assistant Professor of Management Science at the School of Administrative Studies, York University, Canada. He received his PhD in Decision Sciences from NUS Business School, National University of Singapore and did a Post-doctoral fellowship at Carnegie Mellon University, Pittsburgh, USA. He has published papers in refereed Conference Proceedings, Energy, and has also written two book chapters.

1

Introduction

Recent development thinking is that to achieve sustainable development the natural resource base must sustain livelihoods. Energy supply, which is fundamental to any development initiative, is often dependent upon the available resources. Therefore, the link between the energy policies and sustainable development is very imminent. The goal of energy policies for a country should not be confined to meet the increasing demand for energy but also to promote the sustainable use of the available natural resources. The analysis of economy-wide energy decisions requires in the first place an integrated modelling approach, which allows an adequate representation of the interplay between energy, economy and the environment. By and large, nonlinear general

Copyright © 2005 Inderscience Enterprises Ltd.

2

H. Qudrat-Ullah

equilibrium economic approaches have been used to serve the planning needs of energy policy decision makers. Although these approaches intend to provide a picture of long-term trends in energy-economy development, as equilibrium calculations are made on a term-by-term basis, they do not address short-term effects or transitional policy issues (Dyner and Bun, 1997). Moreover, with the ongoing restructuring and liberalisation of economies of many developing countries, the use of empirical modelling and economic approaches is relatively limited in their practical implications. Researchers, especially from system dynamics community, have found system dynamics modelling to be useful in providing some broad insights into the dynamics of energy, economy and environment. For instance, system dynamics models have been developed and applied to national energy policy evaluation (Ford, 1983; Naill, 1992); investments and uncertainty (Ford, 1985), conservation policy analysis (Ford and Bull, 1989); effects of agents on utility performance (Geraphty and Lyneis, 1985); inter-fuel substitution in OECD-European electricity production (Moxnes, 1990); privatisation of electricity industry (Bunn and Larsen, 1992; Bunn et al., 1997); energy efficiency and electricity substitution by gas in the residential and industrial sectors (Dyner et al., 1995). MDESRAP (Model for Dynamics of Electricity Supply, Resources and Pollution) is a dynamic simulation model based on system dynamics methodology (Forrestor, 1961). This model has been applied to energy policy assessment and design both in academic and practitioners’ settings (Qudrat-Ullah, 1999; Qudrat-Ullah and Davidsen, 2001). The purpose of this paper is to describe the development of MDESRAP in terms of motivation, mathematical modelling, and validation process. We will also revisit the applications of MDESRAP to better understand and improve, within the environmental and resource constraints, the energy policy design for electricity supply system of Pakistan.

2

MDESRAP development

2.1 Motivation for MDESRAP In 1990–1991, the government of Pakistan introduced reforms in the electric power supply sector. Lucrative incentives including •

the government’s guarantee for power purchases for WAPDA and KESC (both state-owned units)

•

fuel supply to the electricity producers from public sector entities

•

reduced local earning investments requirements

•

simplified procedures

•

attractive bulk power tariffs (6.5 cents per kilo watt-hour).

In response to these incentives, most of the independent power producers (IPPs) offers included oil, coal, and/or gas power plants. The hydroelectric generation, despite its rich resource base in the country, did not receive any offer. However, on the one hand, Pakistan already imports oil equal to 40% of its need. On the other hand oil, coal and/or gas power plants pose serious challenges to the environment (in the form of greenhouse

MDESRAP: a model for understanding the dynamics of electricity supply

3

gas emissions). Against the environmental emissions, the Kyoto protocols stresses to take into account the greenhouse gas emissions while making energy policies. In this context, the case of Pakistan, where over 30% of the total electricity generation depends upon fossil fuels, provided a rational basis for making an assessment of the existing electricity policy and hence the development of MDESRAP (Qudrat-Ullah, 1999).

2.2 MDESRAP architecture and mathematical formulations The model is organised into seven sectors namely, electricity demand, investment, capital, resource, production, environment, and costs and pricing sectors, as shown in Figure 1. The influence of one sector to another leads automatically to the closure of feedback loops that govern the behaviour of the system. Figure 1

Sectorial interactions of the model

2.2.1 Demand sector The demand sector describes how the electricity demand (ED) is generated based on GDP and the average electricity intensity (EIavg) of GDP. GDP is exogenous to the model. In the model, the average electricity intensity is captured as an exponential smoothing of the short-term indicated electricity intensity (EIsti) over a period (Tstip). ED = EIavg × GDP

(1)

(d/dt)EIavg = (EIsti – EIavg)/Tstip, EIavg(t0) = EIsti(t0)

(2)

4

H. Qudrat-Ullah

The short-term indicated electricity intensity is represented by a co-flow structure as depicted in Figure 2.1 It is influenced by the physics of the electricity generating capital. Figure 2

System dynamics co-flow structure of short-term indicated electricity intensity

Accordingly, the short-term indicated electricity intensity increases when the capital is increased and it decreases when the capital depreciates. The input to the electricity intensity stream is the long-term indicated electricity intensity (LongTIndcElecIntensity)2 while the capital under construction (ElecCaptCons) is the input to the capital. The reference electricity intensity and the effect of electricity price on electricity intensity determine the long-term indicated electricity intensity. The reference electricity intensity represents our case specific base year (1980) value. The effect of electricity price on electricity intensity is dependent upon •

the effect of electricity price on demand

•

the average electricity price

•

the reference electricity price.

If the price of electricity decreases (relative to the reference electricity price), the electricity intensity exhibits a growth pattern. Conversely, the electricity consumption declines when the price rises.

2.2.2 Investment sector The investment sector represents how the investments in electricity capital are made across the generating technologies (coal, oil, gas, hydro-power plants), based on the costs of these technologies, environment premium (due to CO2 tax), investment incentive premium, and resource import dependency premium. The output of this sector is the actual investments being made for the capital of each of the electricity generating technologies. IPPs investments in each of the electricity generating capitals (KI) are represented as an adjustment of indicated investments in electricity generating capital (IKI) over the adjustment period (Tadj). The indicated investments are determined by the total demand of electricity generating capital (TDK), unit capital price (PK) and the share (ST) of each of the competing.

MDESRAP: a model for understanding the dynamics of electricity supply (d/dt)KLi = (IKLi – KLi)/Tadj,

KLi(t0) = IKLi(t0),

IKLi = TDK × PKi × STi

i = coal,gas,oil, hydro

5 (3) (4)

The share of each competing electricity generating technology in the new capital investments being made is represented by a lagged version of the share planned (SPT). The average delay time (CT) represents the construction time of new power plants. Thus, we assume that the choice of electricity generating technology cannot be altered after the plant construction has begun. (d/dt)STi = (SPTi – STi)/CTi, STi(t0) = SPTi(t0)

(5)

Numerous factors are likely to influence the share planned for each technology when the investments decision is being made (Moxnes, 1990). However, we focus, in addition to the standard cost elements, on the investment incentive premium (IIP), the import dependency premium (IDP), the environment premium (ENP) for each technology, and we lump the rest in a power plant technology choice premium (TCP). The standard cost element includes capital costs (CC), (Doperating costs (OC), and fuel costs (FC). Once the total cost for each technology (TC) is determined, we apply a multinomial logit (MNL) model to obtain the share planned of each technology (SPT). We have built this modified total cost structure, with the inclusion of the three kinds of premiums, based upon the work of Moxnes (1990). TCi = CCi + OCi + FCi + IIPi + IDPi + ENPi + TCP

(6)

PSTi = exp (−α × TCi ) / ∑ exp (−α × TCi )

(7)

i

The MNL has only one parameter α (distribution parameter). When α takes an extreme value, then the aggregate choice of the whole sector (all the IPPs) for a technology mimics to the choice of an individual IPP. The lower values of α represent greater variation between the individual choices and the aggregate (sector) choice. When the total costs of all the technologies are equal, the market share is split into equal shares.

2.2.3 Capital sector The capital sector of the model captures the acquisition of production capital, based on capital demand and supply (investments being realised) for each of the technologies. In the model, the electricity (generating) capital means the physical machinery, buildings, equipment, etc. used to produce electricity from fuel (resource). The capital stock (K) is increased by the capital acquisition rate (Kacq) and decreased by the capital depreciation rate (Kdep). Capital is initialised at its equilibrium level. Capital depreciation is proportional to the capital stock. The average life of capital (KT) varies across the technologies. Acquisitions of capital are proportional to the level of capital under construction (KUC). Different construction delays (CT) are assumed depending upon the type of power plant technology. K(acq)i = KUCi/CTi, K(dep)i = Ki/KTi

(8)

Capital stock under construction increases by the new investments made by the IPPs (KI) and decreases when the construction of the power plants is completed. In the absence of demand for new capital, the construction rate reflects the compensation for the

6

H. Qudrat-Ullah

depreciated capital. The delays involved in the construction of each power plant impact the speed at which the capital is accumulated to form a complete plant and this phenomenon is explicitly modelled in this sector. (d/dt)Ki = K(acq)i – K(dep)i (d/dt)KUCi = (KLi/PKi) – K(acq)i

(9) (10)

2.2.4 Resource sector The electricity demand for each technology (EDT) and the safety margin (short-term coverage (SM)) for the on-site resource availability generates the desired demand for each type of the resource (DINV). Until indigenous depletion sets in, the desired resource inventory should be (is) met by the indigenous available resource stock (AINV). The indigenous resource supply (IRS) is determined by the ratio of indigenously available inventory to the desired inventory. DINVi = EDTi × SMi

(11)

IRSi = AINVi/DINVi

(12)

The on-site inventory (OINV) can be filled up to the resource storage capacity (SC). The storage capacity is built based on resource inventory demand. The building of storage capacity takes place with a construction delay (CCT). The local resource (INVadjrate) and the imported resource (IMRrate) add, while the resource consumption (RDEPrate) depletes the on-site inventory stock. At first, the on-site inventory is made up from the local resource. When the indigenous resource (AINV) is depleted, the on-site resource inventory is made up with the imported resource. These imports determine the import dependency (ID). Based on the desired dependency level (DID), a premium (IDP) is charged to gain a dependency reduction. (d/dt)AINVi = –INV(adjrate)i

(13)

(d/dt)OINVi = INV(adjrate)i + IMR(rate)i – RDEP(rate)i

(14)

The premium, as shown in Figure 3, is also one of the results of our study. We have identified this premium as to imply that there is a preference for the least import dependent technology. We consider this finding reasonable. Figure 3

Import dependency premium

MDESRAP: a model for understanding the dynamics of electricity supply

7

2.2.5 Production sector The electricity production (EPRODrate) for each of the electricity generating technologies is determined, based upon the capacity utilisation of the power plants (CU) and the total electricity production capacity of each of the power plants (PC). In our model, the production of electricity is driven by demand, which exerts its influence through the capacity utilisation of the power plants. The electricity production capacity is constrained by the resource potential (RP) and the capital potential (CP), which are the constituents of the total electricity production capacity. EPROD(rate)i = PCi × CUi

(15)

PCi = MIN(RPi, RPi)

(16)

The resource potential to produce electricity (RP) is determined based upon the onsite available stock of fuel (OINV) to be burnt/exhausted, the fuel consumption ratio (CR), the short-term inventory coverage (SM), and the fuel efficiency (FE). RP = [OINVi/(CRi × SMi)]/FEi

(17)

CPi = Ki × KPi

(18)

The electricity generating potential (CP) of each of the technologies is dependent upon the available electricity generating capital (K) and the productivity of this capital (KP). The production potential is proportional to the electricity generating capital as the unit capital productivity is assumed to be fixed for each technology. Consequently, we assume there is no technology (efficiency) improvement of significance. The capacity utilisation of each of the power plants (CU) is a function of the ratio between the demand for electricity (EDT) and the capacity to produce electricity (PC). If this demand/supply ratio exceeds unity, then the capacity utilisation function allows the power plant to run at 100% capacity. But when this ratio drops below unity, the capacity utilisation is reduced in proportion to demand. CUi = IF(EDTi/PCi > = 1,1, IF(EDTi = 0,0,EDTi/PCi))

(19)

2.2.6 Environment sector The environment sector describes how the environment premium ($/MWh) is determined for each of the electricity generating technologies, based on the amount of electricity produced, CO2 emission intensity of fuel, and CO2 tax rates, while taking into account the reference/desired CO2 emission limit. Electricity generation (ElecProdRate) causes CO2 emissions (CO2 EmissionRate). Depending on the volume of the emissions to be mitigated, a tax is charged. We have assumed that the more CO2 a technology emits, the heavier it will be taxed relative to other technologies. This reflects the fact that mitigation costs for a higher level of CO2 emissions is higher as well. A margin for future mitigation investments is also included in the tax rate. The CO2 tax income is invested, over a period of time, to mitigate the emissions. However, not all of the emitted CO2 is treated/contained (sequestered). Some amount of emitted CO2 escapes (CO2Escaped) beyond the national boundaries. This escaped CO2 is a major concern in international mitigation negotiations. The co-flow structure given in Figure 4 helps us keep track of the amount of electricity in both the streams: one for which mitigation investments are being made and the other containing escaped CO2. No investments are made for the

8

H. Qudrat-Ullah

escaped CO2. Also, the average electricity related CO2 intensity (AvgCO2Intensity) is calculated through this structure. The larger the thermal electricity generation, the more CO2 is emitted into the environment. Figure 4

System dynamics co-flow structure of treated and non-treated electricity

2.2.7 Financial sector The financial sector deals with financial conditions, such as costs, prices and net income as they develop as a result of investments. The price of electricity (PE) is the smoothed value of the required price, over an adjustment period of two years. The required price (RP) is the price required to cover the unit cost of electricity production (UC) with a certain gross margin (GM). The unit cost of electricity production consists of the sum of fuel cost (FC) and the operating costs (OC) together with unit CO2 cleanup cost (UCC) if desired (SW1). CO2 tax is assumed to be equal to unit cleanup cost. The gross margin is assumed to cover all of the capital costs (including the fixed operating and maintenance costs), the target rate of return on capital investments and the increased marginal cost due to the capacity constraints. The fuel cost is the reference resource price, modified by the market conditions. RPi = UCi + GMi

(20)

UCi = FCi + OPi + (UCC × SW1)

(21)

The net present value of investments (NPVI) is also determined in this sector. It depends on the net income (NI) of the electricity generating technology, the average physical life of the capital (KT), and the interest rate (IR) associated with competitive investments. The net income is the difference between the Government purchase price of electricity (GPP) and the required price of electricity (RP), determined endogenously based on the cost of electricity production. The relative NPV of investment in a technology determines the attractiveness of the investment incentives. NPVIi = [(1 + IR) ^ (KT)i −1)/(1 + IR ^ [(KT)i×IR] )] × NIi

(22)

MDESRAP: a model for understanding the dynamics of electricity supply

9

2.3 MDESRAP validation and calibration Since system dynamics models are casual models, a model must generate the ‘right behaviour for the right reasons’TM (Barlas, 1996, p.186). The model structure can be compared to descriptive knowledge of real-system structure (known as structural validity); and model behaviour may be compared to observe real-system behaviour (termed as behavioural validity) (Barlas, 1989; Forrester and Senge, 1980; Richardson and Pugh III, 1981; Sterman, 1984). MDESRAP has been exposed extensively to various tests for its structural as well as behavioural validity (for details please see, Qudrat-Ullah, 2004). MDESRAP has been calibrated to represent the electricity generation history (from 1980 to 2000) of Pakistan. The model is initialised so as to be in a steady state under the influence of the exogenous economic conditions (i.e., GDP of Pakistan), with the base year 1980.

3

MDESRAP applications

3.1 Policy assessment MDESRAP has successfully been applied to assess the government of Pakistan’s 1990–1991 energy policy. This policy is described in Table 1 as the base-case scenario. The long-term impact of this policy in a three-dimensional context: Table 1

Scenario descriptions

Scenario

Main features

A: base case

The government’s guarantee for payment of WAPDA and KESC power purchase obligations for 30 years The government’s guarantee for fuel supply from public sector entities Attractive bulk power tariffs (6.5 cents per kilowatt-hour)

B: environmentoriented

In Addition to the base case scenario features, this scenario has:

C: market-oriented

This scenario is basically the same as the base case scenario, but it applies the market price rate (instead of 6.5 cents/kWh) as the government purchase rate and also includes 5% improvement in the generation efficiency

D: self-oriented

This scenario mainly is a reaction to the decreased dependence on domestic resources for electricity generation. The policies included in this scenario are:

Restriction of CO2 emissions to 1990 level More efficient power generation (5% improvement)

Reduced dependence on fuel imports (not more than 5% of the total generation) Efficient power generation (5% improvement) Guaranteed bulk power purchase @ 6.5 cents/kWh for 30 years Guaranteed fuel supply from local resources

10

H. Qudrat-Ullah

•

how does it affect electricity supply

•

how does it influence resource import dependencies

•

how does it contribute to the evolution of, electricity-related, CO2 emissions is assessed.

The impact of 1990–1991 policy reforms resulted in •

a major shift from coal and hydro power plants to oil and gas power plants

•

6,287 millions barrels, of the cumulative oil imports for electricity generation, a 40% above the pre-policy imports

•

the cumulative CO2 emissions level that is 48% above the pre-policy case (Qudrat-Ullah and Davidsen, 2001).

3.2 The alternative policy design MDESRAP has been used to carry out comparative policy impact assessments, with the objective of identifying alternative policies, which have environmental and economic benefits under the constraints of available indigenous resources (Glatzel, 2001). In all four scenarios namely, a base case scenario, an environment-oriented scenario, a market-oriented scenario and a self-oriented scenario as described in Table 1, were analysed. Each scenario covered a 31-year period from 2000 to 2030. The overall comparison of the solutions obtained with the model, in the context of •

losses in GDP (relative to base case)

•

oil import dependency

•

CO2 emissions, is presented in Tables 2–4, respectively.

Table 2

Losses in GDP (%)

Year/scenario

2000

2005

2010

2015

2020

2025

2030 –

A

–

–

–

–

–

–

B

0.25

0.11

0.26

1.62

0.26

12.78

24.32

C

0.25

0.11

0.26

0.52

0.26

1.72

2.45

D

0.25

0.11

0.26

–0.02

0.26

0.25

0.13

Table 3

Oil imports (10 million barrels)

Year/scenario

2000

2005

2010

2015

2020

2025

2030

A

770.7

1,965.1

1,965.1

2,793.1

1,965.1

4,914.1

6,287.0

B

762.1

1,564.0

1,564.0

1,854.5

1,564.0

2,478.7

2,839.1

C

762.1

1,940.0

1,940.0

2,758.4

1,940.0

4,858.4

6,158.6

D

762.1

1,921.3

1,921.3

2,718.8

1,921.3

4,747.4

5,992.0

MDESRAP: a model for understanding the dynamics of electricity supply Table 4

11

CO2 emission (10 million tons)

Year/scenario

2000

2005

2010

2015

2020

A

70.98

108.54

147.18

185.19

147.18

B

70.98

62.04

17.01

16.69

17.01

16.41

40.56

C

70.98

108.49

146.50

182.86

146.50

260.67

308.33

D

70.98

108.55

185.12

210.01

185.12

269.09

321.86

Main conclusions drawn (Qudrat-Ullah, 2004):

from

the

comparative

scenario

2025

2030

269.36

322.28

analyses

were

•

In terms of economic growth, no conclusive scenario emerges. In all scenarios, the economy adjusts itself to slower growth with smaller GDP values. The losses in GDP in each scenario are indicative of the costs to the economy in terms of lost output resulting from the corresponding decrease in the electricity intensity of GDP.

•

In terms of oil import dependency, it seems quite possible to keep the pace of the economy at levels similar to the base levels but with a significant reduction in fuel import dependency (via scenario D).

•

In terms of environmental emissions, the regulatory structures, as in scenario B, might help to carve environment-friendly energy policies and simultaneously support sustainable use of available, but limited, indigenous resources of nations.

4

Conclusion

In this paper, we have described the motivation, development, validation and usage of a dynamic simulation model: MDESRAP, capable of capturing the dynamics of underling sectors of electricity supply system. The government of Pakistan’s reforms in the electric power supply sector, introduced in 1990–1991, provided the context for MDESRAP development. MDESRAP is based on system dynamics approach. The main strength of MDESRAP lies in its ability to represent the interactions among •

patterns of energy demand

•

substitution between electricity and its competitive alternative

•

investments in electricity supply sector

•

construction of new plants and depreciation of older plants

•

capacity constraints

•

emission constraints

•

indigenous resource constraints

•

financial structure

•

the regulatory processes.

12

H. Qudrat-Ullah

We have also shown how MDESRAP has been used to policy assessment and policy design domains. Also, MDESRAP is equally adaptable to a wide variety of problems within the domain of electricity supply system.

References Barlas, Y. (1989) ‘Multiple tests for validation of system dynamics type of simulation models’, European Journal of Operation Research, Vol. 42, pp.59–87. Barlas, Y. (1996) ‘Formal aspects of model validity and validation in system dynamics’, System Dynamics Review, Vol. 12, No. 3, pp.183–210. Bunn, D. and Larsen, E. (1992) ‘Sensitivity reserve margin to factors influencing investment behaviour in the electricity market of England and Wales’, Energy Policy, Vol. 29, pp.420–429. Bunn, D., Larsen, E. and Vlahos, K. (1997) ‘Complementary modelling approaches for analyzing several effects of privatization’, in Bunn, D.W. and Larsen, E.R. (Eds.): Systems Modelling for Energy Policy, John Wiley, Chichester, pp.303–325. Dyner, I. and Bunn, D. (1997) ‘A systems simulation platform to support energy policy in Columbia’, in Bunn, D.W. and Dyner, I. (Eds.): Systems Modelling for Energy Policy, John Wiley, Chichester, pp.259–271. Dyner, I., Smith, R. and Pena, G. (1995) ‘System dynamics modelling for energy efficiency analysis and management’, Journal of Operational Research, Vol. 46, No. 10, pp.1163–1173. Ford, A. (1983) ‘Using simulation for policy evaluation in the electric utility industry’, Simulation, pp.85–92. Ford, A. (1985) ‘Short lead time technologies as a defense against demand uncertainty’, in Plummer, J., Oatman, E. and Gupta, P. (Eds.): Strategic Management and Planning for Electric Utilities, Prentice Hall, Englewood Cliffs, NJ. Ford, A. and Bull, M. (1989) ‘Using system dynamics for conservation policy analysis in the Pacific Northwest’, System Dynamics Review, Vol. 15, No. 1, pp.1–16. Forrester, J. (1961) Industrial Dynamics, Productivity Press, Cambridge, MA. Forrester, J. and Senge, P. (1980) ‘Tests for building confidence in system dynamics models’, TIMS Studies in Management Sciences, Vol. 14, pp.209–228. Geraphty, D.M. and Lyneis, J. (1985) ‘Feedback loops: the effect of external agents on utility performance’, in Plummer, J., Oatman, E. and Gupta, P. (Eds.): Strategic Management and Planning for Electric Utilities, Prentice Hall, Englewood Cliffs, NJ. Glatzel, W. (2001) ‘Climate technologies: opportunities for leap-forging in developing countries’, Proceedings of the OECD Seoul Conference November 2000, OCED Publications, Paris, pp.161–171. Moxnes, E. (1990) ‘Interfuel substitution in OECD-European electricity production’, System Dynamics Review, Vol. 6, No. 1, pp.44–65. Naill, R.F. (1992) ‘A system dynamics model for national energy policy planning™’, System Dynamics Review, Vol. 8, No. 1, pp.1–19. Qudrat-Ullah, H. (1999) Dynamics of Electricity Supply, Resources, and Pollution, M. Phil. Thesis, Department of Information Science, University of Bergen, Bergen, Norway. Qudrat-Ullah, H. (2004) ‘Resources, pollution, and development of sustainable energy policies’, in Quaddus, M.A. and Siddique, M.A.B. (Eds.): A Handbook of Sustainable Development Planning: Studies in Modelling and Decision Support, Edward Edgar (to be available in October 2004), UK. Qudrat-Ullah, H. and Davidsen, P. (2001) ‘Understanding the dynamics of electricity supply, resources, and pollution: Pakistan’s case’, Energy, Vol. 26, No. 6, pp.595–606.

MDESRAP: a model for understanding the dynamics of electricity supply

13

Richardson, G.P. and Pugh III, A. (1981) Introduction to System Dynamics Modeling with Dynamo, Productivity Press, Portland, OR. Sterman, J.D. (1984) ‘Appropriate summary statistics for evaluating the historical fit of system dynamics models’, DYNAMICA, Vol. 10, No. 2, pp.51–66.

Notes 1

MDESRAP structure and the simulations analysis are carried out using the software ‘Powersim Constructor’ Version 2.5d, by Powersim AS. PO Box 206, Isdalsto, Norway. 2 Nomenclature is given in Appendix.

Appendix: Nomenclature ActualInvst

Actual investments [$/year]

AvgTimeForInvst

Average time for investments [year]

AvgLifeElecCapt

Average life of electricity capital [year]

AvgTimeForRain

Average time for rainfall [year]

AvgIntensityUnderCons

Average intensity under construction [kWh/(103 × $)]

CapBuildTime

Fuel storage capacity build-up time [year]

CapIntens

Capital intensity [kWh/(103 × $ × MW)]

CapIntensCon

Capital intensity under construction [kWh/(103 × $ × MW)]

CapIntensCon_Rate

Capital intensity construction rate [(kWhxYear)/(103 × $ × MW)]

CapIntens_AcqRate

Capital intensity acquisition rate [(kWh × year)/(103 × $ × MW)]

CapIntensDepr

Capital intensity depreciation rate [(kWh × year)/(103 × $xMW)]

CO2Escaped

Escaped CO2 (emissions gone outside the national boundary) [ton (CO2)]

ConsDel

Power plant construction delay [year]

CumAnnualProd

Cumulative annual production (of electricity) [MWh]

EffectOfElecPriceOnDemand

Effect of electricity price on its demand [dimensionless]

ElecCapt

Electricity capital [MW]

ElecCaptCons

Electricity capital construction (rate) [MW/year]

ElecCaptAcq

Electricity capital acquisition (rate) [MW/year]

ElecCaptDepr

Electricity capital depreciation (rate) [MW/year]

ElecCaptUndCons

Electricity capital under construction [MW/year]

ElecCaptProdctivity

Electricity capital productivity [MWh/year-MW]

FuelConsumpRatio

Fuel consumption ratio [BTU/MWh]

FuelEfficiency

Fuel efficiency [dimensionless]

FOM_Cost

Fixed operating and maintenance costs [$/MWy]

GovPurchasePriceOfElec

Government purchase price of electricity [$/MWh]

14

H. Qudrat-Ullah

IndustryRate

Industry rate [%/year]

InvAdjstTime

Inventory adjustment time [year]

NonTreatedElec

Non-treated electricity (related emissions are not treated) [MWh]

TreatedRate

Treated electricity rate (related emissions treatment rate) [MWh]

OpertCost

Operating costs [$/MWh]

OrigIndigAvailRes

Original indigenous available resources [BTU]

OtherPrem

Other premium [$/MWh]

PriceOfElecCapt

Price of electricity capital [$/MW]

RefElecIntensity

Reference electricity intensity [kWh/$]

RefResourcePrice

Reference resource price [$/MWh]

ResourcePriceElasticity

Resource price elasticity [dimensionless]

Short_Term_Coverage

Short-term coverage [BTU]

TargetRetnOnInvest

Target return on investments [%/year]

TimeToAdjstPriceOfElec

Time to adjust the price of electricity [year]

TimeToAdjstElecCapt

Time to adjust electricity capital [year]

TimeToRealiseChng

Time to realise change [year]

α

Coefficient for the distribution of investments [kWh/$]