JET PROPULSION LABORATORY California Institute of Technology • 4800 Oak Grove Drive, Pasadena, California 91103 January 16, 1978 Distribution: Enclosed is a copy of the A. D. Little report entitled "Impacts and Benefits of a Satellite Power System on the Electric Utility Industry." The report is sent to you because of your known interest in the subject matter. This document was prepared as one of a series of topical reports on the JPL Study of Satellite Power System Impacts and Benefits. The study was sponsored by the NASA Office of Energy Programs (OEP) and was performed under the program management of Mr. Simon V, Manson of the OEP Solar Energy Division. The Study Task Manager at JPL was Mr. Irving Stein of the JPL Control and Energy Conversion Divisiono Other reports of this series will be sent to you as they become available. Marshall E. Alper, Manager Solar Energy Program
DISTRIBUTION NASA HEADQUARTERS H. Calahan (2) Code NT D. Cauffman Code ST J. Disher Code MT S. Fordyce Code ECF H. Fosque Code TA T. Hagler Code MTE R. Johnson Code MTE R. LaRock Code NT S. Manson (5) Code NT S. Sadin Code RX E. Schmerling Code ST S. Tilford Code SU D. Winter (2) Code SB DEPARTMENT OF ENERGY Solar and Geothermal F. Koomanoff (5) Code DSE J. Madewell Code DSE H. Marvin Code DSE E. Willis Code AIR Office of Assistant Secretary for Environment R. Blaunstein (6) Code DTO M. Minthorn, Jr. Code DBER Argonne National Laboratory 9700 South Cass Avenue Argonne, IL 60439 Attn: T. Wolsko (6) Argonne National Laboratory 400 N. Capitol Street NW Washington, DC 20001 Attn: C. Sandahi Battelle-PNL Box 999 Richland, WA 99352 Attn: W. Bair (5) Los Alamos Scientific Laboratory Box 1663 Los Alamos, NM 87545 Attn: J. Hopkins (3) Planning Research Corporation 7600 Old Springhouse Road McLean, VA 22101 Attn: C. Bain (4) Solar Energy Research Institute 1536 Cole Boulevard Golden, CO 80401 Attn: P. Rappaport JOHNSON SPACE CENTER G. Arndt Code EJ5 L. Bell (3) Code EZ C. Covington (4) Code EW4 H. Davis Code ER 0. Garriott (6) Code SA J. Loftus Code A R. Piland (2) Code EA4
MARSHALL SPACE FLIGHT CENTER K. Fikes (15) Code PD11 C. Guttman (5) Code PS04 J. Murphy Code PA01 G. Von Tiesenhausen Code PS01 LEWIS RESEARCH CENTER J. Ward (2) Code 49-3 AMES RESEARCH CENTER B. Newsom Code 239-1 Dr. J. Sharp Code 200-7 P. Sebesta Code 239-1 JPL INTERNAL DISTRIBUTION M. Alper 169-527 J. Bowyer 125-159 S. Brunstein 114-B13E R. Caputo 277-202 H. Cotrill, Jr. 198-102 K. Dawson 198-102 R. Dickinson 238-528 D. Dipprey 198-102 R. Forney 169-422 E. Framan 169-527 R. Goldstein 122-123 R. Hortop 238-528 M. Lavin 156-203 F. Livingston 79-6 R. O'Toole 79-200 D. Ross 198-220 E. Sheldon 198-220 I. Stein 198-220 R. Stephenson 169-420 V. Truscello 277-202 P. Wiener 198-220 EXTERNAL DISTRIBUTION Aerospace Corporation Los Angeles, CA Attn: M. Wolfe Arthur D. Little, Inc. Cambridge, MA Attn: P. Glaser Boeing Aerospace Seattle, WA Attn: G. Woodcock (2) ECON Princeton, NJ Attn: G. Hazelrigg Grumman Aircraft Corporation Bethpage, NY Attn: J. Mackovciak, Jr. (2) Raytheon Corporation Sudbury, MA Attn: 0. Maynard Rockwell International Space Division Seal Beach, CA Attn: G. Hanley (2)
IMPACTS AND BENEFITS OF A SATELLITE POWER SYSTEM ON THE ELECTRIC UTILITY INDUSTRY Final Report for Jet Propulsion Laboratory California Institute of Technology Pasadena, California Contract No. 954639 C-80020 Arthur D Little, Inc
Arthur D.Little.Inc. ACORN PARK- CAMBRIDGE,MA.02140- (617) 864-5770-TELEX 921436 14 July 1977 This report is submitted by Arthur D. Little, Inc. (ADL) in fulfillment of JPL Contract No. 954639. The ADL study manager was Dr. B.M. Winer. The work presented herein was performed by a study team which included Mr. Gerald Larocque, Dr. James Nicol, Dr. Peter Glaser and Dr. John Bzura. In the course of this program, the staff has received informatJon and advice from several sources. We are grateful for the information provided by Dr. Piet Bos of EPRI. We are particularly grateful to Mr. Irving Stein, the Program Technical Manager for the Jet Propulsion Laboratory. Dr. B.M. Winer Senior Staff Member Physical Systems Research Section CAMBRIDGE. MASSACHUSETTS ATHENS BRUSSELS LONDON PARIS RIO DE JANEIRO SAN FRANCISCO TORONTO WASHINGTON WIESBADEN
TABLE OF CONTENTS List of Figures iii List of Tables v 1.0 INTRODUCTION AND SUMMARY 1-1 1.1 Executive Summary 1-1 1.2 Reliability and Stability 1-5 1.3 Possible Ownership of the SPS 1-15 1.4 Utility Participation in SPS Related RD&D 1-19 1.5 Utility Liabilities Associated with the SPS 1-19 1.6 Structure of the Report 1-20 2.0 RELIABILITY AND STABILITY 2-1 2.1 Background 2-1 2.2 Power System Stability Characteristics 2-6 2.2.1 Introduction and Results 2-6 2.2.2 System Dynamics 2-7 2.2.3 Protection Devices 2-12 2.2.4 Northeast Blackout 2-15 2.3 Power Pool Reliability 2-19 2.3.1 Introduction and Summary 2-19 2.3.2 Formulation of the Problem 2-30 184.108.40.206 Definitions 2-30 220.127.116.11 Power Pool Loads 2-34 18.104.22.168 Number of Required Generators 2-37
3.0 POSSIBLE OWNERSHIP OF SPS 3-1 3.1 Introduction 3-1 3.1.1 Summary 3-1 3.1.2 General Financial Characteristics of the Generation Mix 3-5 3.2 Purchase of the SPS by a Utility or Consortium of Utilities 3-14 3.3 “Leasing" the SPS 3-21 3.4 Pricing SPS Produced Energy at the Incremental Cost of Alternative Generation 3-23 3.4.1 Introduction 3-23 3.4.2 Calculated Maximum Discount Rates 3-24 4.0 UTILITY PARTICIPATION IN THE SPS RD&D PROGRAMS 4-1 5.0 UTILITY LIABILITY DUE TO THE ADVERSE EFFECTS OF SPS RELATED ACTIVITIES 5-1 APPENDIX A - CALCULATIONS OF THE POWER POOL GENERATING MARGINS REQUIRED TO MEET THE LOLP CRITERIA A-l APPENDIX B - CHANGE IN POWER POOL COSTS DUE TO SPS B-l APPENDIX C - CASH FLOW ANALYSIS - SPS ENERGY PRICED AT THE COST OF ALTERNATIVE BASE LOAD GENERATION C-l
1.1 Required Percent Installed Margin as a Function of the Power Pool Size - Power Pools Containing One SPS 1-12 1.2 Required Percent Installed Margin as a Function of the Power Pool Size - Power Pools Containing Two SPSs 1-13 1.3 Required Percent Installed Margin as a Function of the Power Pool Size - Power Pools Containing Six SPSs 1-14 2.1 National Electric Reliability Council 2-4 2.2 Frequency Distribution as a Function of Time After Perturbation 2-10 2.3 Transmitted Power as a Function of the Generator Power Angle 2-13 2.4 Diagram of Load Control Areas and Power System Interconnections, CANUSE and PJM 2-16 2.5 Demand for Power in the Two Power Consuming Elements of the Composite Power Pool as a Function of the Time-of-Day 2-26 2.6 Required Percent Installed Margin as a Function of the Power Pool Size - Power Pools Containing One SPS 2-27 2.7 Required Percent Installed Margin as a Function of the Power Pool Size - Power Pools Containing Two SPSs 2-28 2.8 Required Percent Installed Margin as a Function of the Power Pool Size - Power Pools Containing Six SPSs 2-29 2.9 Demand for Power From Conventional Generators in the West Coast Component of the Composite Power Pool as a Function of Time 2-38 LIST OF FIGURES
3.1 Demand for Electric Power Over a Weekly Cycle 3-6 3.2 Power Demand as a Function of Time 5% Growth Rate 3-8 3.3 Power Demand as a Function of Time 7% Growth Rate 3-9 3.4 Average Fuel, Fixed and Total Per Unit Costs 3-18 3.5 Average Per Unit Fuel Costs, Fixed Costs and Total Costs 3-19 3.6 A Total Debt as a Function of Time 3-30 3.6B Debt Incurred Yearly as a Function of Time 3-30 3.7 A Total Debt as a Function of Time 3-31 3.7B Debt Incurred Yearly as a Function of Time 3-31
LIST OF TABLES 1.1 Installed Generating Margin (GWe) For the Various Pools as a Function of the Circumstances 1-10 2.1 Regional Load Density (1974) 2-3 2.2 Required Installed Generating Margin (GWe) For a Range of Power Pools According to Various Circumstances 2-24 3.1 Incremental Costs of Conventional Generation 3-25 3.2 Maximum Allowed Discount Rate as a Function of Inflation Revenues Set Equal to Fuel Costs of Nuclear Generator 3-26 3.3 Maximum Allowed Discount Rate as a Function of Inflation Revenues Set Equal to Fuel Costs of Coal Generators 3-26 3.4 Maximum Allowed Discount Rate as a Function of Inflation Revenues Set Equal to Fuel Costs of Oil Generators 3-27 3.5 Revenues Set Equal to Fuel Costs of Nuclear Generator at Maximum Allowable Discount Rates 3-28 3.6 Revenues Set Equal to Fuel Costs of Coal Generator at Maximum Allowable Discount Rates 3-29 3.7 Revenues Set Equal to Fuel Costs of Oil Generator at Maximum Allowable Discount Rates 3-29
A.l Length of Time (Hours) During Which Power Demand is Between m and m-1 GWe 30 GWe Power Pool A-8 A.2 Length of Time (Hours) During Which Power Demand is Between m and m-1 GWe 40 GWe Power Pool A-9 A.3 Length of Time (Hours) During Which Power Demand is Between m and m-1 GWe 50 GWe Power Pool A-10 A.4 Length of Time (Hours) During Which Power Demand is Between m and m-1 MWe Composite (30 GWe + 30 GWe) Power Pool A-12 A.5 Required Number of Conventional Generators in a 30 GWe Power Pool as a Function of Maintenance Interval and Circumstances A-14 A.6 Required Number of Conventional Generators in a 40 GWe Power Pool as a Function of Maintenance Interval and Circumstances A-15 A.7 Required Number of Conventional Generators in a 50 GWe Power Pool as a Function of Maintenance Interval and Circumstances A-16 A.8 Required Number of Conventional Generators in Each Portion of the Composite Power Pool (30 GWe and 30 GWe) as a Function of Maintenance Interval and Circumstances A-21
INTRODUCTION AND SUMMARY 1.1 Executive Summary The purpose of this limited study was to investigate six specific issues associated with interfacing a Satellite Power System (5 GW) with large (by present standards) terrestrial power pools to a depth sufficient to determine if certain interface problems and/or benefits exist and what future studies of these problems are required. The issues investigated and the conclusions reached are as follows: 1. Stability of Power Pools Containing a 5 GWe SPS Using present control methods, the power pools investigated in this study are unlikely to be able to maintain stable operation without shedding part of the load if the SPS shuts down unexpectedly. This might be a severe problem and further studies of (a) the likely magnitude of the problem, (b) the most co?t effective method of alleviating the problem are needed. 2. Extra Reserve Margin Required to Maintain the Reliability of Power Pools Containing a 5 GWe SPS The use of any type (SPS or conventional) of 5 GWe generator instead of five 1 GWe generators requires a significant increase in the power pool reserve margin if the system reliability is to be maintained; the cost of the extra capacity need not be excessively expensive. The problem is significant and deserves further study, but a solution is available at a reasonable cost. 3. Use of the SPS in Load Following Service (i.e. in two independent pools whose times of peak demand differ by three hours) The use of the SPS in this manner does not allow the economics of the SPS to be directly compared with the
economics of terrestrial peaking plants. The use of the SPS reduces the magnitude of the peak demand for conventional generation capacity in each pool by only 2% but reduces the duration of this peak significantly. The effect would be to change the optimum mix (base, cycling and peaking capacity) of generation equipment in the pools. Further study of this issue is required before any further conclusions can be reached. 4. Ownership of the SPS and Its Effect on SPS Usage and Utility Costs Of the three ownership and energy marketing alternatives considered, the most promising appears to be ownership of SPS by an independent corporation, not the operating utility, and the sale of energy generated by the SPS under long-term contracts. 5. Utility Sharing of SPS related RD&D Costs A review of the electric utilities' financial commitment to EPRI indicates that, given the most optimistic assumptions about the desire of the utilities to support SPS related RD&D, the utilities will be unable to contribute any more than 10% of the required $44 billion. Present utility and EPRI RD&D funding priorities indicate that the electric utilities will be unwilling to contribute as much as 1% of the SPS's development costs. 6. Utility Liability for SPS Related Hazards At present, the magnitude and geographic limits of the potential hazards are poorly defined. No utility can afford to assume the legal liabilities which might be associated with these risks.
Other conclusions reached in this study are as follows: • The large size and high plant cost of the SPS are major impediments to its inclusion in terrestrial power pools as presently constituted. • SPS outages which are limited to the actual duration of an eclipse of the sun by the earth would have no effect on the power pool's fixed costs (total required amount of generating capacity), if the power demand in :the pool varies by a factor of two during the day. • The large size of the SPS will probably force the power pool to "shed load" if and when the SPS shuts down unexpectedly; this could be true even if there were enough spinning reserve available to compensate for the loss of the generation capacity. • Utility ownership of the SPS will be financially difficult if the "fuel adjustment clause" continues in widespread use. • The risks associated with selling SPS energy at the incremental costs of terrestrial base-load alternatives are probably too large to be assumed by a private corporation.
Recommendations Because of the limited resources available for this study, models with insufficient detail to fully validate the conclusions had to be used. The following, more extensive studies of the SPS- utility interface are recommended before any final decision is made to build the SPS. • Perform a stability analysis for a specific large power pool to determine (1) the required stability of the SPS output, and (2) the probable "loss of load" associated with an unexpected SPS shutdown. • Investigate various methods and the associated costs of reducing SPS induced stability problems, e.g., transmitting SPS power via multiple high voltage de transmission lines (1 GW per circuit) to five different power pools remote from the rectenna site. • Calculate the optimum generating mix and operating costs for each of the two separate power pools in which the SPS is used in load following service. • Re-calculate the reserve margin requirements of the power pool with and without the SPS using more realistic models of the power pool generation mix and the SPS. • Calculate the cost of the required increase in the power pool spinning reserve caused by the inclusion of the SPS.
• Calculate the power pool operating costs with and without the SPS using a more realistic model of the power pool (use Production Costing Programs). • Using utility expansion planning programs and a more realistic model of the power pool than was used in this study, calculate the utility costs (fixed and operating) as a function of the year after the SPS becomes operational, if (a) the utilities purchase the SPS, (b) the utilities purchase energy from the owner of the SPS, or (c) they follow normal (non-SPS) expansion. • Determine how the availability of SPS power is likely to affect the utility generation expansion plans. • Determine the maximum amount of SPS power that can be absorbed by power pools of various sizes. • Perform those studies which will will be required to define the magnitude and location of the hazards, if any, likely to be associated with the SPS. 1.2 Reliability and Stability The overall reliability of the bulk electric power network has been given the highest priority by the utilities and the FPC. The ★ following are some of the many aspects of system reliability. * "Design of Electric Power Systems for Maximum Service Reliability" by C. Concordia, CIGRE> 1968, Report No. 32-08.
• The assurance of sufficient generating and transmission capacity, in view of the projected loads and equipment availability, so that the Loss of Load Probability (LOLP) shall not exceed the design level; • The ability to withstand the sudden loss of a major generator or transmission line, without inducing any other outages; • The ability to withstand line faults without losing any generators; • The minimization of system breakdown, as measured by loss of generation, cascading line outage, and loss of load when disturbances more severe than expected may occur; and • The ability to restore service quickly and smoothly after a complete system breakdown and source interruption. Probability that power demand exceeds generation capacity.
It should be noted that roughly half of these system aspects relate to the ability of the system to respond to disturbances without undue reaction (what we shall call system stability) and the other half refers to the adequacy of generation and transmission equipment to meet the demand for electric power (what we shall call system reliability). These two criteria are related; a system which is inadequate to meet the power demand is more likely to over-react to certain types of system disturbances. The question addressed in this report was: What kind of stability and reliability problems will arise when an SPS is added to a power pool? Within the limits of available resources, the purpose of the study was to describe the nature of the problems and estimate their magnitudes. The problems investigated were: • Stability • Frequency disturbances caused by sudden changes in the amount of generation capacity in the power pool. • Effect of protection device operation on machine stability. • Reliability • Reserve margin requirements to maintain prescribed reliability in power pools containing one or more SPS with a variety of assumed outage characteristics. • Use of the SPS in load following service.
The relatively qualitative investigation of stability indicates that: 1. The sudden loss of the 5 GWe SPS output would probably cause a loss of load whenever the power pool was meeting a total load of roughly 40 GWe or less. The largest power pool considered in this study (peak demand = 50 GWe) meets a load of 40 GWe or less 88% of the time. 2. Sudden fluctuations in the SPS output could cause the operation of protective devices which themselves could exacerbate the stability problems. The investigation of reliability turned out to be basically a calculation of the total required installed capacity needed in a power pool if one or more SPS's (each with a generating capacity of 5 GWe) were installed instead of a number of conventional generating plants (each with a generation capacity of 1 GWe). This analysis was concerned primarily with the size of the proposed SPS, and therefore, most of the results would apply equally well to a 5 GWe terrestrial plant. The results indicate that whenever a 5 GWe generation is used instead of five 1 GWe generators (no change in the forced outage rate) an additional one to two gigawatts ($124 - $250 million) of extra reserve capacity (gas turbines at $125/kW) must be added if the system reliability is to be maintained. The magnitude of the assumed reliability criterion is not critical, since it is not likely to be changed when the SPS is added to tne power pool. The total amount of reserve generating capacity required in various power pools was calculated for power pools having yearly peak power demands of either: • 30 GWe, or • 40 GWe, or • 50 GWe, or a
• Composite Power Pool made up of two independent 30 GWe Power Pools whose times of peak demand differ by 3 hours. These power pools contained either • No SPS (all conventional equipment), or • One (5 GWe) SPS, or • Two (5 GWe) SPS, or • Six (5 GWe) SPS. Three different scheduled interruptions of the power from the SPS were considered: • Power interruption due to eclipses only during the actual eclipse period; no scheduled maintenance requirements. [This was a best case calculation.] • Power interruption due to eclipses only during the actual eclipse period, plus scheduled maintenance for 20% of the year. [This was a worst case calculation.] • Power interruption due to eclipses for the entire day for all days during which an eclipse occurs (90 days). [This was a worst case calculation; the SPS is unlikely to be economically attractive under these circumstances.] The magnitude of the installed reserve under each of the indicated conditions is entered in Table 1.1. The difference between the entry of interest and the entry for the power pool which does not contain an SPS is the extra installed margin that is required by the
TABLE 1,1 Installed Generating Margin (GWq) For the Various Pools as a Function of the Circumstances
SPS. For example: If a power pool, which has a peak power demand of 50 GWe contains no SPS, only 10 to 11 GWe of installed margin (60 to 61 GWe total) are required to provide for system reliability. If this same power pool contains an SPS which must be shut down for scheduled maintenance, 12 to 13 GWe of installed margin are required. The power pool which contains an SPS needing scheduled maintenance requires two more gigawatts of generating capacity than does the power pool that contains no SPS. If the SPS needs no scheduled maintenance, only one more gigawatt of generating capacity would probably be needed (11- 12 GWe minus 10-11 GWe). The results shown in Table 1.1 indicate that if one or more 5 GWe generators (SPS, nuclear or fossil fuel) are installed in a power pool, the installed generating margin must be increased if the system reliability is to be maintained. The percentage increase would depend on the size of the power pool; the larger the power pool, the smaller the required percentage increase. To demonstrate how the installed margin must vary with the power pool size, the percentage installed margin is plotted as a function of the power pool size in Figures 1.1, 1.2 and 1.3. * The plotted values for the composite power pool clearly indicate that the composite power pool cannot be treated as if it were a 60 GWe power pool. The additional generating capacity that the results of this study indicate will be required need not be expensive. The extra capacity will not be used very often and could be in the form if inexpensive peaking units ($125/kW), causing an increased capital require- ** ment of $250 million, 3.3% of the cost of the SPS ($7.6 billion) and Two independent 30 GWe Power Pools whose times of peak demand differ by three hours. "Space-Based Solar Power Conversion and Delivery Systems Study — Interim Summary Report" by ECON, Inc., March 1976, Report No. 76- 145-IB.
FIGURE 1.1 REQUIRED PERCENT INSTALLED MARGIN AS A FUNCTION OF THE POWER POOL SIZE POWER POOLS CONTAINING ONE SPS
FIGURE 1.2 REQUIRED PERCENT INSTALLED MARGIN AS A FUNCTION OF THE POWER POOL SIZE POWER POOLS CONTAINING TWO SPSs
FIGURE 1.3 REQUIRED PERCENT INSTALLED MARGIN AS A FUNCTION OF THE POWER POOL SIZE POWER POOLS CONTAINING SIX SPSs
roughly $50 million/year for fuel. If a completely redundant antenna were built to eliminate the need for scheduled maintenance, the total cost increase (including 1 GW of gas turbines) would be $1.47 billion, 19% of the cost of the SPS. An additional conclusion was reached while actually performing the calculations; if the SPS is shut down by the earth eclipses for only the duration of the eclipse, the eclipses will have no effect on the system reliability. The demand for power during these eclipse periods was only half the daily peak and the probability that other generation would not be available to supply the required power was virtually zero. If the shutdown were to last from one hour before the eclipse to one hour after the eclipse, the results would be the same. This particular problem had no effect on the system LOLP and should be considered further only if it is expected that the daily load curve was tending to become flat. The composite power pool was found to be unaffected by either the SPS maintenance requirements or problems due to the eclipse. Because the power produced by satellite in this power pool could be used in some way or other throughout the year, it is understandable that the maintenance requirements of the ground stations would have little effect on the installed margin. The margin's insensitivity to the eclipse stems from the large size of the required margin when the pool contains no SPS and the uncertainties in the calculation. 1.3 Possible Ownership of the SPS Three different ownership and/or energy pricing arrangements for the SPS have been investigated. These arrangements were: • Purchase of the SPS by a utility or consortium of utilities.
• Purchase of the SPS by an independent corporation and ''lease” (commitment to purchase a share of the SPS energy) of the output by several utilities. • Purchase of the SPS by an independent corporation and the energy sold to the utilities, at below cost initially, at a price equal to the incremental cost of the utilities' most expensive base load generator. How the SPS is purchased and by whom can determine how it is used. Of these three arrangements, the most promising appears to be the purchase of the SPS by an independent entity (corporate or governmental) and "lease" of the output by several utilities. While all the calculations performed in this analysis assumed that the capital cost of the SPS was $7.6 billion, the general conclusions reached can be used to infer the effect of the more recent, significantly higher estimate of $12.2 billion. The basic conclusion of this study, i.e., that the "leasing" arrangement is the most promising of the three arrangements considered, would be true if the higher cost had been assumed. The results of this investigation are as follows: 1. Utility Ownership of the SPS • When the ($7.6 billion) SPS first becomes operational, a very small increase in the total cost of meeting the demand for electrical energy will probably occur.
• If the capital cost of the SPS is $12.2 billion, the inclusion of the SPS related costs in the utility rate structure would require an increase in the total cost of electrical energy to the consumer. • Utilities which use a semi-automatic fuel adjustment rate to recoup the cost of fuel will have to request a sizable increase in their base rates to cover their increased plant equity when the SPS comes on-line. Fuel rate reductions can occur within a month; base rate increases can take as long as a year to obtain. The higher the capital cost of the SPS, the greater may be the financial stress caused by regulatory delays. 2. "Leasing" of the SPS Output by the Utilities • The cost of purchasing energy could be recouped by many utilities via fuel adjustment rates. • At present, the reduction of the utility capital requirements caused by "leasing" energy from the SPS would have a beneficial effect on the utilities' financial ratings. It is not clear that this situation will prevail over the next fifty years, nor is it clear if the utilities would accept this arrangement over such a long term. • Since the utilities make no profit on purchased energy, the effect of the SPS on the total cost of electrical energy would be the same for both the utility ownership and the private ownership/utility leasing plans (assuming that the discount rate is the same for both the utility and the private corporation).
3. SPS Energy Sold at the Incremental Cost of Base-Load Alternatives • If the inflation rate continues at roughly the same as present rates, it would be possible to price energy from an SPS (capital cost = $7.6 billion) at the incremental cost of alternative fossil fueled generation and eventually make a profit. The size of the profit depends on the inflation rates. • If the capital cost of the SPS is significantly higher than $7.6 billion, the inflation rates necessary to eventually make a profit using this pricing alternative, would be significantly greater than the present inflation rates. • Pricing SPS generated energy in this manner requires the operation of the SPS at a loss for roughly twenty years. The risks associated with this arrangement are too large for private industry- financial guarantees from the government would be required. • If the government provides financial guarantees to a corporation intending to price SPS energy in this manner, the interpretation of this decision may be that either the government is willing to subsidize the SPS or that the government expects the inflation rate to continue at its present level or higher.
1.4 Utility Participation in SPS Related RD&D While the participation of the electric utilities in the SPS research, design and development (RD&D) program may be desirable, utility activities in this area are likely to be very limited during the next five years. EPRI's budget for all solar energy research during this time period is only 2% of EPRI's total budget. The total research EPRI budget for the next five years is roughly $1 billion, including an allowance for inflation. Of this, only $20 million (approximately $4 million/year) has been allocated for all forms of solar energy research, including solar heating and cooling. Unless EPRI's priorities shift significantly, the funding available from this source to support SPS related R&D will probably be small. Even if EPRI supported SPS-related RD&D at the same rate as all other solar energy projects combined, its contribution between now and 1995 would probably be less than 1% of the required total of $44 billion. If all of EPRI's resources were devoted to the SPS, EPRI could only contribute roughly 10% of the $44 billion required. The probability of attracting substantial participation by individual utilities in SPS related research is also small; utility research priorities are primarily near-term and investment in the SPS is unlikely to be a high priority item. 1.5 Utility Liabilities Associated with the SPS Whoever owns the SPS - the electric utilities, a private or semi- private corporation or a government agency, this owner could be liable for all the adverse effects that could results from SPS related activities; the cost of these liabilities would presumably be added to the cost of SPS generated electrical energy via the cost of insurance. At present, too little is known about the potential adverse effects either to:
• identify all the possible liabilities, • estimate the magnitude of all identified liabilities, • reliably estimate the cost of meeting the liabilities, or • determine whether the electric utilities would assume these liabilities. In the past, the electric utilities have assumed the liabilities associated with the degradation of radio and television reception along transmission right-of-ways. This liability is localized geographically and can be reasonably well defined before the transmission circuit is energized. On the other hand, the similar problem associated with the interference of the SPS microwave beams with communications channels, radar, etc., may be neither localized geographically nor well defined before the first two SPSs are built. The utilities would be unlikely to accept this type of liability as a condition of purchasing an SPS or SPS delivered energy. 1.6 Structure of the Report Each of the six issues investigated in this study is discussed in some detail in the following chapters. Since there was some relationship among the first three issues, they were grouped together in Chapter 2. All others are described in independent chapters. The results of the study in each area are summarized at the beginning of each chapter so that each chapter can stand alone.
2.0 RELIABILITY AND STABILITY 2.1 Background The overall reliability of the bulk electric power network has been given the highest priority by the utilities and the FPC. The * following are some of the many aspects of system reliability. • The assurance of sufficient generating and transmission capacity, in view of the projected loads and equipment availability, so that the Loss of * A Load Probability (LOLP) shall not exceed the design level; • The reliable operation of the individual pieces of equipment; • The ability to withstand the sudden loss of a major generator or transmission line, without inducing any other outages; • The ability to withstand line faults without forcing any generators to shut down; • The minimization of system breakdown, as measured by loss of generation, cascading line outage, and loss of load when disturbances more severe than expected may occur; and a "Design of Electric Power Systems for Maximum Service Reliability" by C. Concordia, CIGRE, 1968, Report No. 32-08. Probability that power demand exceeds generation capacity.
• The ability to restore service quickly and smoothly in case of a partial or complete system breakdown and source interruption. It should be noted that half of these system aspects relate to the ability of the system to respond to disturbances without undue reaction (what we shall call system stability) and the other half refer to the adequacy of generation and transmission equipment to meet the demand for electric power. These two criteria are related; the less the excess of generation capacity over power demand, the more likely is the system to react with instability to certain types of system disturbances. The question addressed in this report was how is the SPS likely to affect either the stability or reliability of the existing or expected power pools? The resources allocated for this study were too small to allow an evaluation of these problems in the depth they deserve. The purpose of the study was to describe the nature of the problems and to estimate their magnitudes. Regional Reliability Councils The 1965 "Northeast Blackout", followed by another extensive blackout in another area in 1967, had wide repercussions within the industry. Many questions were raised such as: • Are the planning criteria correct? • Are design concepts adequate? • Should interconnections between power systems be strengthened or eliminated? Extensive studies of these questions were undertaken by both the utilities and the Federal Power Commission (FPC). The
results of these studies indicated a need for a high degree of coordination of the system planning, design, and operating functions between interconnected utilities. The National Electric Reliability Council (NERC) and the Regional Reliability Councils were established to encourage this coordination. The nine Regional Reliability Councils encompass essentially all of the power systems of the United States and the Canadian systems in Ontario, British Columbia, Manitoba, and New Brunswick. The area covered by each of these councils and the abbreviations commonly used for each are shown in Figure 2.1. Each of the Regional Reliability Councils has developed slightly different reliability criteria for testing and evaluating simulated future system designs which reflect the differences which exist in geography, population density, load pattern, power sources, etc. The variation of the load densities from region to region is shown in Table 2.1 as an example. However, the overall goals of the various councils are essentially uniform. Regional boundaries are only arbitrary lines of demarcation, thus criteria in adjoining regions or continguous utilities on regional borders must be compatible. Joint agreements between regions exist and studies to assure compatibility of reliability criteria are performed. Table 2.1 Regional Load Density (1974) (contiguous U.S. only)
The North American bulk power supply is not only the largest but, by far, the most reliable electrical network in the world. The 1975 NERC annual report stated: "The record of the past year (1975) attests to the successful operation of the network even under various stresses caused by violent weather conditions, equipment failures and several acts of sabotage." Another mute testimony to the strength of the system was provided by its successful operation during the adverse conditions caused by the fuel shortages and bad weather conditions of January 1977. The Reliability Councils and the operating utilities and/ or power pools are quite different. Each of the Reliability Councils is based on a voluntary agreement among the member utilities to uphold the basic principles of reliable system planning and operation; membership in the Reliability Council is a voluntary agreement. An operating utility is a centrally controlled organization having the direct responsibility of building, operating and maintaining the equipment (generation, transmission and distribution) necessary to meet the load in its area reliably and at the lowest possible cost. An operating power pool centrally controls all the generation and transmission equipment owned by its member utilities; the contracts which define the power pool contain legal penalties for nonconformance to reliability criteria. A decision to build and operate a 5 GWe SPS to be placed in one of the Reliability Regions may have a significant effect on the regional planning process; the effect may be no greater than the effect of placing any similarly sized generator in the region. The purpose of this section of the report is to investigate the likely magnitude of the effects. Since each reliability council operates somewhat differently, it has been impossible to do more than indicate the circumstances under which problems would occur so as to guide the SPS design team in their efforts.
2,2 Power System Stability Characteristics 2.2.1 Introduction and Results Predicting the stability of a large scale power network is an extremely complex problem. In general, because of the intricate interactions among the various lines, generators and load devices, a full modelling of an electric power network requires the solution of a complex system of coupled time varying differential equations. Solutions, generally, cannot be obtained within normal time and budget constraints on a digital computer. They are certainly beyond the resources of this limited study, but even with the larger studies one must usually be content with results based on average network properties and with qualitative descriptions of potential difficulties at the level of individual elements. This section presents a qualitative discussion of the system characteristics in order to convey an appreciation of the problems that can occur. It should be noted that the stability characteristics discussed herein, are the same as those required of conventional generation capacity. The results of this relatively qualitative investigation of stability indicate that: The sudden, unexpected loss of the SPS output would cause a loss of load whenever the power pool was meeting a total load of roughly 40 GWe or less. The largest power pool considered in this study meets a load of 40 GWe or less, 88% of the time. 2. Sudden fluctuations in the SPS output could cause the operation of protective devices which themselves could exacerbate the stability problems. The key points to be made in the following discussion are that if satellite power systems create frequent fluctuations in
system generation capacity, the introduction of such a power source may increase the number of transients of the power network and cause frequent redistributions of power flow throughout the network. The SPS should be designed so that any fluctuations in output power occur as slowly as is necessary to allow earth-bound regulator systems to correct for them without creating significant transients. Section 2.2.2 provides a discussion of the transient in system frequency due to system dynamics resulting from a loss of generation capacity. This transient is of concern because off-frequency operation has a severely adverse effect on many types of load elements and also places undue stress upon generator turbines as a result of governor operation at other than design frequencies. In Section 2.2.3 the effects of protection devices operation on machine stability is discussed, indicating the potential for large scale network instability as a result of switching operations. An example of stability problems is found in the Northeast blackout where a variation in the load caused a normally functioning protection device to initiate a sequence of events resulting in loss of power to most of the northeastern United States. This incident is discussed in some detail in Section 2.2.4. 2.2.2 System Dynamics After the loss of a generator unit in an electric power network, a frequency transient will occur whose precise characteristics are a function of many factors; e.g., the magnitude of power loss with respect to the remaining generation, the time constants of the remaining generators and the dynamics of the governors attached to the network. The detailed solution for such transient problems is complex, and in most instances, it is possible only to deal with average system properties. In so doing, it is necessary to apply weighting factors to the properties of each of the generators in the network. There are many ways in which these factors may be selected, but the basic analysis is unaltered.
In a system with only a single generator, or in which all generators are identical, the average system frequency is governed by the following differential equation: The same equation gives a good approximation to the solution for a more complex system if parameters derived from appropriate weighted averages of shaft kinetic energy, governor dynamics, etc., are used. Solving this equation for £(t) assuming that there is a change in the available generation capacity at t = 0 provides the following expression for the transient response:
where uj and co. are the complex natural frequencies of the system given o 1 as the roots of the equation: The meaning of these expressions can be demonstrated if values typical of a network whose generators are primarily steam turbines are substituted for the system parameters. The form of the transient is shown in Figure 2.2; the maximum deviation from nominal frequency is given approximately by
FIGURE 2.2 FREQUENCY DISTRIBUTION AS A FUNCTION OF TIME AFTER PERTURBATION
△P is the fraction of the total power being generated, g The above result indicates that a sudden reduction in generation capacity (assuming the system is able to absorb the loss of generation with available spinning reserve) will create an approximately sinusoidal frequency transient whose peak value is directly proportional to the magnitude of the power loss; this helps to define the required stability of the output of the satellite power system. If for example, a 2% change in frequency is the maximum to be tolerated, the satellite power system would have to maintain its generation level so as to produce maximum power fluctuations of no more than .12 of the total network generation at the time of the change in the SPS output. If the total power pool demand were 30 GWe, the maximum allowable fluctuation would be 3.6 GWe. If the total power pool demand were 10 GWe, the maximum allowable fluctuation would be 1.2 GWe. Normally, a power pool will have sufficient generation capacity on-line to meet the expected load plus a certain amount of spinning reserve; the required amount of spinning reserve is equal to either a percentage of the maximum expected load (typically 3-7% of the system load) or to the output of the largest generator on-line, which ever is larger. This ensures that the system will be able to absorb any unexpected loss of generation without large frequency changes. The large size of the SPS will probably require a significant increase in the level of spinning reserve and the operating cost of the power pool would consequently increase. The modern use of load shedding relays have reduced the probability of large scale system shutdowns occurring as a result of the sudden loss of generation capacity. These relays disconnect part of the load so that the system can still meet the larger part of the load. Even if the spinning reserve were provided for the example given, the sudden loss of the SPS output would force a loss of load operation of the relays whenever the total load is less than 42 GW.
This loss of load is undesirable except as an alternative to the total shutdown of the power pool. 2.2.3 Protection Devices The use of circuit breakers to protect lines against faults is common practice even though their use may cause generator instabilities. This section describes the nature of the instability that can be caused by the normal operation of circuit breakers. When a line is faulted, generators connected electrically close to the fault experience a sharp decrease in their load (since the voltage at the fault is zero, no real power can flow in the faulted line except for line loss) while other units in the system are required to pick-up the fraction of load isolated from the generators on the other side of the fault. This means that during the faulted condition, some generator rotors are accelerated while others are decelerated. Consequently, when the fault is cleared, the system is in a configuration in which some generators are advanced and some are retarded from their previous equilibrium values. There is a maximum angular displacement from which a generator can recover a stable equilibrium. To illustrate this point, consider the simplified case of a generator supplying an infinite bus through a series of transmission lines. Under such circumstances, the power balance of the system is described by the following differential equation.
The maximum power that can be transferred is sinusoidal with respect to power angle. For two different circuit configurations, the maximum power transfer as a function of power angle might appear as curves I and II in Figure 2.3. The difference might be a higher reactance between the generator and the infinite bus (e.g., switching out of a line) in curve II. FIGURE 2.3 TRANSMITTED POWER AS A FUNCTION OF THE GENERATOR POWER ANGLE In condition I, the equilibrium value of 6 is When the line is switched out, the generator rotor begins to accelerate because the power transmitted is less than the mechanical power to the rotor. The rate of change of the rotor angle is given by
The integral is graphically represented by the difference between areas and A£ on Figure 2.3. — will be zero when = A^. At this point, where the electrical power is greater than the mechanical power, and the rate of change of 6 reverses, the rotor swings back towards angle 6^. Because there are always losses to damp a real system, the rotor will eventually stabilize at a new equilibrium angle $2* Generator instabilities can occur because there is a critical value 6^ for 6. If the rotor exceeds this critical power angle, the generator cannot regain equilibrium. This critical power angle exists because, as shown on the figure, has a maximum value equal to the area between curve II and the line P = P . If A- is larger than this maximum, the rate of change of 6 never reaches zero, and the power balance tends further to increase the machine's angle. Thus, there are certain critical machine angles which must not be exceeded during a switching operation or else some of the machines will not be able to reestablish equilibrium states. If the power network is subjected to frequent changes in generation capacity, the power distribution over the lines of the network will be changing often. It is not inconceivable that redistribution of power over a network due to generation fluctuation could cause the power on some line to exceed the setting of its protective device, causing the line to be disconnected, creating the sort of transient problem described above in addition to the frequency transient set up by the loss of generation. Since the switching of the line again redistributes the power flow, a chain reaction could occur, magnifying the stability problem. The magnitudes and frequencies of fluctuations likely to initiate a chain reaction of this sort are difficult to forecast; the sort of system breakup just discussed is a line-by-line and machine-bymachine process which does not lend itself to description by average characteristics. It is more than a simple cascade of analyses like that
of the previous two pages, because after the first event the system is usually not in equilibrium when the next discontinuity occurs. Determination of such a sequence of events requires a detailed transient load flow analysis at each change of network configuration (i.e., loss of generation, switching of lines, or change of load) coupled with a line-by-line examination of power flow and protection device setting, along with examination of machine stability limits at the power demands involved. This sort of analysis is tantamount to a complete simulation of the entire power network. In a study of this sort, it is impossible to make general statements about the magnitude and frequency of power shifts likely to cause large scale network shutdown. However, the potential for such situations does exist and the larger and more frequent the power fluctuations, the greater the probability of such an occurrence. 2,2.4 Northeast Blackout The Northeast Blackout is an example of instability problems which arose from the normal operation of protective devices. Before discussing the series of events leading to the blackout, it is necessary vo indicate some of the important characteristics of the Canada-United States Interconnection (CANUSE). Hydroelectric power constituted approximately 26 percent of the CANUSE generation and is largely concentrated in the Niagara Falls area. Most of this power is transmitted to loads located far from the generation site. Power which is generated by Power Authority of the State of New York (PASNY) plants in the Niagara Falls area is transmitted in large part by twin 345 kV lines from Niagara to Albany to New York City. Niagara and PASNY were interconnected with the Connecticut Valley Electric Exchange (CONVEX) and the New England Electric System (NEES) by one 345 kV line, one 230 kV and five 115 kV lines (see Figure 2.4.) Seven transmission lines carrying from 115 to 230 kV connect CANUSE with the Pennsylvania-New Jersey-Maryland (PJM) power pool.
FIGURE 2.4 DIAGRAM OF LOAD CONTROL AREAS AND POWER SYSTEM INTERCONNECTIONS, CANUSE AND PJM.
The event leading to the Northeast Blackout originated at the Sir Adam Beck generation complex at Kingston, Ontario (part of the Ontario Hydro System). Immediately prior to the blackout, the Hydroelectric Power Commission of Ontario was meeting a system load of approximately 6400 MW with Sir Adam Beck generating 1335 MW and with a 500 MW inflow on two tie lines with PASNY. Approximately 200 MW of the 500 MW inflow was being returned to New York via other interconnections. The Beck complex is connected with the Toronto load center via five parallel 230 kV lines. In 1963, a backup relay on one of the 230 kV lines had been set substantially below the line's rating at 375 MW in order to achieve coordination with other protection devices in the power network. On the day of the blackout, the average power flow in this line reached a level of 365 MW and at 5:16 PM, the 375 MW rating was exceeded during a fluctuation in load. This caused the line to be opened by the protective relay, resulting in the power flow to Toronto being distributed among the remaining four lines, causing each of them to be overloaded with the result that they were disconnected by their relays. Thus, within a few seconds, the 1335 MW being generated at Sir Adam Beck was isolated from its load center in Toronto. This caused the generators in the Niagara area to accelerate due to the loss of electric load and with this increase in speed came a rapid increase in power output. This power had to be distributed via the interconnections with PASNY and caused the remaining lines interconnecting Ontario and PASNY to become overloaded. Thus, the sole interconnection between Ontario at New York existed at Niagara where the Beck plant was isolated from Ontario but still connected to New York. The excess power output from the Niagara area could not be handled by the remaining lines and resulted in the stability limit opening of the two 345 kV lines