Volume 6, Number 1, Fall 2005

---

The Dynamics of Transformer Purchasing in a Transformed Electric Market

David I. Eromon, PhD

Department of Electronics, Computer and Information Technology

North Carolina A&T State University

Greensboro, NC 27411

dieromon@ncat.edu

Abstract

The electric utility industry as it has operated over the past 70 years is changing. Historically, most of the industry has been vertically integrated, meaning that one company provided electric services from generation to transmission to distribution to customer service. It has been widely argued that the generation portion is not a natural monopoly and should be separated from the other functions of electric service. Generation would then become a competitive market from which distribution companies, or even retail customers, would purchase their requirements. Transmission would be controlled by a separately-regulated Independent System Operator (ISO). This would help to maintain reliability of the system and avoid the problems of market power in which a company could use its transmission lines to limit competition for generation and increase its prices. The purpose of this paper is to examine the calculations used to evaluate transformer costs and how they may be affected by the restructuring of the industry.

RESTRUCTURING—THE MARKET AS A WHOLE

Although generation will become deregulated, distribution of electricity will continue to be a regulated business. It is a natural monopoly in that it is most cost-effective to have a single set of distribution lines for a given region, rather than multiple companies each having electric wires to ultimate consumers. Because of the deintegration of the industry, the distribution business will look more like current distribution-only utilities. Currently, roughly 14% of the country is served by distribution utilities that have no generation of their own⁴. These companies sign long-term supply contracts with either nearby investor-owned generating utilities or public power suppliers such as TVA. Their costs are based on the terms of these contracts rather than the cost of specific generating plants. Most of these distribution companies are either publicly owned cooperatives or municipal utilities.

In the future, retail customers may have contracts with different generation providers at different prices. Distribution companies will bill consumers for the use of their wires; these bills will include the costs of losses. Customers could pay for the losses by either having their generation supplier provide extra power, or simply pay the distribution company to procure the power.

In this environment, utilities will continue to need to purchase transformers. The purpose of this paper is to examine the calculations used to evaluate transformer costs and how they may be affected by the restructuring of the industry. Section 2 describes what the transformer costs are, section 3 describes the equation used for comparing the total operating costs (TOC) of different transformers with an emphasis on those parameters most likely to change by restructuring, section 4 discusses the incentives for companies to purchase efficient transformers, and section 5 is a summary.

Distribution Transformer Losses

Measurement of Losses

There are two types of losses associated with transformers: core losses (also called no-load losses) and load losses. Core losses occur at a constant value all the time that the transformer is energized. They are due to the resistance of Load losses resistance which vary as a square of the energy flow across the transformer. Consequently, they will be much higher proportionately during peak loads than during off-peak times.^3,5,7

A distribution utility does not specifically measure its transformer losses. Instead, it meters how much electricity enters and exits its system. The difference in the amount measured is due to a combination of line losses, transformer losses, internal use by the utility, and unmetered theft. Some large transformers may have measurements taken around them that allow the losses to be calculated, but for the most part, utilities rely on the manufacturers’ specifications to determine transformer losses. Unless the power flow through the transformer as a function of time is measured, the actual losses can only be estimated.

The appropriate cost of the power lost is the cost of the power that the company would not have bought if it did not have the losses. This depends on the generation used to provide it and the contracts the utility has in place for purchasing power. Currently, distribution-only utilities have contracts that include both a demand payment (based on peak demand for a given period) and an energy payment (based on the total energy used over the period.) Rates may vary based on the time the energy is used and the overall quantity.^1,2 Other charges also enter into the total bill to reflect ancillary services provided by the generation and transmission provider. Utilities can also buy blocks of power on the wholesale market under various terms and conditions.

Under restructuring, a distribution company will continue to need a contract for provision of power, both for itself and for those customers who choose to continue to buy from the company. And, unless the utility can physically disconnect the customers, it will also have to provide power to those customers with supply contracts from other companies if their suppliers are not producing at the time of demand. The power the distribution company needs may be purchased through contracts with generators, power brokers, from a spot market, or a combination of these. The utility may use financial mechanisms such as futures contracts or options to hedge on the risks of price changes. These varied markets greatly complicate determining the cost of lost power because a determination must be made on which supply was used to provide the lost energy. As a simplification, the utility may designate one of its sources, such as the spot market, as the marginal supply.

Cost Recovery and Incentives for Efficiency

For billing purposes, transformer load losses are combined with other line losses, the energy used by the utility, and losses due to theft. Losses specific to transformers are not measured. The Public Utility Commission determines a percentage markup to bills based on historical values of the total loss and internal use. These factors then increase the rates for customers. Load and no-load transformer losses (explained below) are not priced separately and not at the marginal cost of the power lost.

Why will distribution utilities want to buy more efficient transformers? The utility is allowed to pass on the cost of new capital equipment through their rates and even earn a reasonable profit on their investment. While this theoretically makes utilities neutral to investment (and some would argue encourages investment), at the actual decision-making level there is a tendency to purchase the item with the lowest first cost. Capital budgets are often set early, and managers attempt to maximize the equipment purchased. The idea behind the TOC equation is to factor in the long-term costs (such as losses) not included in the capital budget but real to the utility nevertheless.^7,8,9 The costs must be put on the same basis as the up-front purchase cost, taking into account other costs, such as taxes and return on investment, that the utility actually pays for an asset.

The long-term cost of losses is paid by the utility through the difference in measured power into the system and out through the individual meters. Because transformer losses are not specifically measured, the utility uses a mark-up of power rates to account for this loss plus other unmeasured uses within the utility and theft. With regulation, if the losses are reduced, then regulators will lower this percentage mark-up and pass the savings on to consumers. There is consequently little incentive for the utility to reduce losses, except that in between rate hearings they are able to keep the savings as extra profits (regulatory lag).

While the companies may continue to be regulated, they might not use the traditional cost-plus form of rate making. Instead, many may use performance-based ratemaking (PBR). Under PBR, prices are set based on both the utility’s costs and certain industry standards for cost or quality. If the utility is able to improve their performance, either by lowering costs or improving services, they are allowed to keep all or a part of the savings. If the utility can lower its power losses below the percentage that is included in rates, it can keep the savings as profits. For example, if the accepted loss rate is five percent and the utility can lower the losses from transformers (or line losses, internal use, or theft) to four percent, then it can retain the extra percent of revenue. Even after rate hearings, if overall prices remain at or below their agreed-upon cap, the utility may continue to receive the extra profits. The philosophy behind this is that letting the utility profit from the cost savings will give them incentives to lower costs more dramatically than under standard rate making. Additional details can be found in A Primer on Incentive Regulation for Electric Utilities⁵.

TOC Calculation Effects from Restructuring

The Total Operation Cost (TOC) equation is used to factor in the long-term costs of a transformer with the up-front cost. It adds to the initial price the cost of the energy lost due to energizing the transformer and passing power through, plus factors in the long-term asset costs of taxes and return on investment.

Equation

According to the Working Group on the Guide for Distribution Loss Evaluation⁶, the equation for the Total Operation Cost of a new transformer is

TOC = Price + Cost of Core Loss + Cost of Load Loss

Cost of Core Loss = A($/Watt) x Core Loss Watts x Loss Multiplier

Cost of Load Loss = B($/Watt) x Load Loss Watts x Loss Multiplier

A and B represent equivalent first cost of the Core (No Load) losses and the Load losses, respectively. They include some simplifying factors to translate the cost of the energy lost into the initial year present values. It also simplifies the analysis by ignoring the long term asset value of the transformer and the consequences to revenues and expenses over time. HPY represents the hours per year that the transformer is energized, typically 8760 hours. Below, we discuss the other factors in the equation that could be affected by restructuring.

Price

The first factor is the price of the transformer. In reality, the utility does not pay just the initial price alone because the transformer is a long-term asset that the utility will have to pay taxes on and from which the utility will earn a return. The long term cost of the transformer will be higher because of these other factors. Rather than multiply the price by the fixed charge rate (FCR) to put all costs on an equal annualized basis, the other two parts of the TOC equation are divided by the FCR, simplifying the equation. The FCR is described later.

Core Loss

Core (or No-Load) Losses represent the energy lost from the transformer even when no power is transferring through it. This amount is constant for all hours that the transformer is energized.

1-System Cost

The first factor of note in the TOC equation with regard to restructuring is the System Cost (SC). This is meant to define the cost to provide the generation and transmission capacity needed to make up the lost power. Historically, this value has represented the annual capacity charge for base load power (because the Core Loss represents a constant, or base, load). Large coal plants or nuclear plants could have a high capacity cost, from $100 to over $400/kW-year. However, more recent applications of the TOC equation recommend using the average for all generation because the power lost is on the margin and so comes from all plants on a time-averaged basis. The amount should roughly equal the demand charge that is charged to large customers by the distributor. Alternatively, the marginal cost of capacity over the life of the transformer should be close to the long-run marginal cost of new capacity. Most projections show that future capacity additions will mainly be gas-fired combustion turbines or combined-cycle units. These have capital costs of around $400/kW according to the Energy Information Administration⁴. Using an FCR of 15% gives an annualized charge of $60/kW.

Under restructuring, it is not clear whether there will be any capacity charge. Instead, utilities or customers would purchase power based solely on the cost of energy used (¢/kWh). More recent deliberations, including the restructuring legislation passed in California, recognize that some type of capacity charge may be needed to ensure that sufficient capacity exists to meet peak demands. Several options have been explored to determine the capacity charge. One method is for the ISO to establish a secondary market for capacity.² The ISO would determine the amount of additional capacity they need to have available for system support (spinning reserve, voltage support, etc.) and accept bids from enough plants to meet their capacity needs. Details must be worked out on a number of issues: whether this payment goes to all suppliers or just those that make capacity available but are not called upon to provide energy, how to incorporate other ancillary services, how to ensure the economically efficient plants are used, determining the time period for each block of the market, and other technical details of the market.

In England, prices are based on both an energy pool that plants bid into and long-term fixed-price contracts between suppliers and the distribution companies⁷. The PJM Interconnection (a power pool in the Pennsylvania, New Jersey, and Maryland area) includes a capacity reservation system in their restructuring proposal in which they calculate at the start of the year a necessary reserve margin and consequent capacity payment to all generators to ensure sufficient capacity availability⁸. In general, it can be expected that the SC for generation will decline greatly from current values, if not totally disappear.

The Transmission component of the SC under restructuring may undergo a similar, although smaller, change. Transmission costs may be bundled into an energy charge, or the distribution company may be assessed a charge based on demand on the transmission system. Since transmission will continue to be regulated, prices will not change as radically as generation prices.

2-Energy Cost

The Energy Cost (EC) is the second factor that could be affected by restructuring. The marginal cost (or price in a deregulated market) of the energy lost will vary over time as demand and supply change. Since the amount lost is constant over time, the cost of the energy lost would be equal to the time-averaged marginal cost of energy, including transmission losses.

3-Fixed Charge Rate

The third factor is the Fixed Charge Rate (FCR). This converts annual costs into a present value based on the cost of capital to the firm, length of time of the investment and tax regulations on the asset. Other assumptions include the tax rates, percentage debt and equity, rates of return, and both book and tax life of the transformer.

As an example, assume a utility spends $1000 on a new transformer (Table 1). In the first year, the utility has expenses of over $187 for the transformer.^{1, 2} This includes $39 for interest, $53 profit to shareholders, $32 for income taxes, $33 for depreciation, and $30 for property taxes. The amount declines over time, reaching zero after year 30. Using the average cost of capital for the utility, the total cost of the transformer is not $1,000 but $3,099.⁵ Discounting this using the average cost of capital for the utility gives a net present value of the transformer of $1,351.

The constant payment that would give the same present value as the actual stream, divided by the original cost, gives the FCR. In our example, an annual payment of $136.10, when discounted, equals $1,351. Dividing by $1,000 gives an FCR of 13.61%. Table 2 shows what happens to the FCR by changing some of the input values. The class life and tax life for distribution equipment are specified in Internal Revenue Service regulations as 30 years and 20 years, respectively¹. Property tax rates can vary across the country from less than 1% to over 10%⁹.

Table 1: Example Calculation of Total Revenue Requirements

	Annualized payment of NPV		$136.1
	Levelized Fixed Charge Rate		13.61%		Capital	Component	Weighted
Price Life (Years) Tax life Income Tax Rate Prop Tax Rate		$1,000		Source	ization	Cost	Cost
	30			Debt	50%	8.0%	4.00%
	20			Preferred	10%	10.0%	1.00%
	38.0%			Common	40%	11.0%	4.40%
	3%			Total Capital	100%		9.40%
		Revenue Requirement
	Average	Interest	Return on	Income	Book	Property
	Rate Base	Payment	Equity	Tax	Deprec.	Tax	Total
1	$981.75	$39.27	$53.01	$32.49	$33.33	$29.50	$187.61
2	933.65	37.35	50.42	30.90	33.33	28.50	180.50
3	887.61	35.50	47.93	29.38	33.33	27.50	173.65
4	843.47	33.74	45.55	27.92	33.33	26.50	167.04
5	801.10	32.04	43.26	26.51	33.33	25.50	160.65
6	760.35	30.41	41.06	25.17	33.33	24.50	154.47
7	721.11	28.84	38.94	23.87	33.33	23.50	148.48
8	683.26	27.33	36.90	22.61	33.33	22.50	142.67
9	645.64	25.83	34.86	21.37	33.33	21.50	136.89
10	608.01	24.32	32.83	20.12	33.33	20.50	131.11
11	570.39	22.82	30.80	18.88	33.33	19.50	125.33
12	532.77	21.31	28.77	17.63	33.33	18.50	119.55
13	495.15	19.81	26.74	16.39	33.33	17.50	113.77
14	457.53	18.30	24.71	15.14	33.33	16.50	107.98
15	419.91	16.80	22.68	13.90	33.33	15.50	102.20
16	382.29	15.29	20.64	12.65	33.33	14.50	96.42
17	344.67	13.79	18.61	11.41	33.33	13.50	90.64
18	307.05	12.28	16.58	10.16	33.33	12.50	84.86
19	269.43	10.78	14.55	8.92	33.33	11.50	79.08
20	231.81	9.27	12.52	7.67	33.33	10.50	73.30
21	202.67	8.11	10.94	6.71	33.33	9.50	68.59
22	182.00	7.28	9.83	6.02	33.33	8.50	64.96
23	161.33	6.45	8.71	5.34	33.33	7.50	61.34
24	140.67	5.63	7.60	4.66	33.33	6.50	57.71
25	120.00	4.80	6.48	3.97	33.33	5.50	54.08
26	99.33	3.97	5.36	3.29	33.33	4.50	50.46
27	78.67	3.15	4.25	2.60	33.33	3.50	46.83
28	58.00	2.32	3.13	1.92	33.33	2.50	43.20
29	37.33	1.49	2.02	1.24	33.33	1.50	39.58
30	16.67	0.67	0.90	0.55	33.33	0.50	35.95
	Total	$519	$701	$429	$1,000	$450	$3,099
	NPV at 9.4%	$255	$345	$211	$331	$209	$1,351

Even with industry restructuring, it is expected that distribution companies will continue to be regulated. Consequently, the parameters involved in the FCR will not change much because of restructuring. Arguments can be made that the distribution company could become a less risky investment because it no longer has power plants in its asset base, which are more risky than the distribution business. However, since many utilities are already distribution only, there should not be a great change in their financial status.

Load Loss represents the energy lost that is dependent on the power actually flowing through the transformer. The amount is not just a linear function of the power flow but a square of the power.

1-System Cost

This factor should be the same as used in the Core Loss calculation. Although some have used the lower capital cost for peaking plants as a measure, future distribution-only utilities should only use a lower factor if their supply contracts provide a differential based on time, which is unlikely. They will be paying the same cost per kW regardless of when that demand occurs.

2-Responsibility Factor

The Peak Responsibility Factor (RF) adjusts the system cost to reflect the proportion of the transformer load that actually contributes to the peak load of the utility as a whole. This factor remains the same with restructuring, although it is less important with the decrease in relative importance of the SC. The importance of the timing of the demand on the transformer versus demands on the system will be captured, instead, in the energy cost as it varies over time and demand.

3-Energy Cost

The TOC equation as currently defined calculates the impact of this disproportionate loss compared to load, but it assumes a constant energy cost (EC) for this loss. Under restructuring, more utilities will go to real-time pricing, in which the price of energy is a function of the overall demand and supply. Market forces will enter into the equation, and prices will be much higher in times of scarcity during peak loads. Since this is also when transformer losses are highest, the equation must more heavily weigh the price of those lost kilowatt-hours, not just the quantity lost, through the LsF factor.⁵ (The pricing of lost energy at the peak is somewhat blurred for two reasons: the peak load on the transformer does not necessarily match the peak for the system, as explained by the RF factor; and prices don’t necessarily match peak demands because of plant outages raising prices at non-peak times.)

While discussions on the TOC recognize that the avoided cost of power should be used, they fail to recognize this compounding of losses at the higher loads and prices. This means that the weighted average price should not be weighted by the energy used but the square of the energy used. For example, Figure 1 shows a typical load function on a transformer, where the average load is only 37% of the peak and a representative price curve for the same time. The time-weighted average price of power is only 2.26 ¢/kWh and is the value for EC used in the Core Loss equation. However, when weighting the power purchased by the square of the power lost, the price paid (EC) becomes 3.29¢/kWh, 45% higher.

Figure 1. A typical Load Function on a Transformer

This value is very sensitive to the time of the peak load. If we shift the peak prices to when the load is lower (such as between the hours 5 and 21), the energy² weighted price drops to 2.39 ¢/kWh. This points out the significance of the timing of the load on the transformer in relation to the price of energy in determining its TOC.^1,2,3 High-efficiency transformers will be most cost-effective if the peak on the transformer (and consequent energy loss) occurs when prices are high. If the load is more constant for the transformer (i.e., a higher load factor), the TOC will be less affected by peak prices. If we raise the load factor to 70% for the example, the energy² weighted price drops to 2.44 ¢/kWh.

Since the EC is to represent the levelized cost of the power lost over the life of the transformer, it becomes necessary to project the price over the next twenty to thirty years. The EIA’s Annual Energy Outlook for 1997 projects a -0.6% growth in prices (in constant $) between 1995 and 2015⁴. Under restructuring, prices could fall even more.

4-Loss Factor

The loss factor (LsF) converts the peak load to the actual energy lost based on the square of the actual load profile. As mentioned above, it does not factor in the price differential for the energy at different load levels.

5-Equivalent Annual Peak Load

The Equivalent Annual Peak Load (PL) levelizes the expected growth in peak demand over the life of the transformer. Under restructuring, peak demands may grow less quickly than overall growth in energy sales. If the market changes to real-time pricing, it can be expected that price sensitive customers will lower their demands during high-priced periods.¹⁰ They may choose to shift their demand to a lower cost time or forego the use of the power altogether. Either way, this will lower the growth of the peak and increase the load factor. This factor will also influence the other factors: RF, SC, EC, and LsF.

SUMMARY

Restructuring will bring great changes to the electric industry as a whole. The major effect on distribution will be the unbundling of this function from the other components, generation, and transmission. The costs of power lost through transformers will be based on market prices rather than the avoided generation from a utility’s own plants. Overall costs will go down, and there will be a shift of relying more on the price of energy rather than demand. Prices will be more volatile, as real-time pricing becomes more prevalent.

The TOC equation will remain the same, but some of the factors will need to be reconsidered. The SC will decline as it will reflect the cost of capacity needed to maintain reliability, instead of all capacity. The EC for the Core Losses will be the time-averaged price of energy, while the EC for Load Losses will need to factor in the different prices when the transformer losses actually occur. The FCR will remain largely the same, although the distribution company may have different capital costs than current integrated utilities. Peak growth may slow, but overall demand may increase with lower prices. This will increase the load factor over time and change the values for RF and LsF.

In a recent article in Public Utilities Fortnightly, George Pleat⁸ describes the benefits of buying energy efficient transformers to a distribution-only utility of the future.

Under the assumption that the state utility commission will regulate the stand-alone distribution company through some kind of performance-based or price-capping mechanism, Disco [distribution company] management should acquire an incentive to purchase the transformer with the lowest life-cycle cost. Purchasing higher-cost but more-efficient transformers with reasonable payback periods (maybe 10-15 years) should help reduce long-run operating costs, boosting profits for stockholders.

More specifically, the incremental cost associated with a more expensive but more efficient transformer is recovered through savings associated with reduced losses (calculated on marginal costs). It would behoove the Disco to pursue these type of decisions.⁸

Despite restructuring, distribution companies will continue to be regulated. Where current ratemaking tends to make a utility neutral to purchasing high efficiency transformers, performance-based ratemaking will build in incentives for them to purchase more efficient transformers, if cost-effective.⁸ The TOC calculation, when used with the proper values for parameters, should show the lowest-cost, and highest-profit, transformers for the utility to buy.

REFERENCES

[1] Commerce Clearing House, Inc., 1993, 1993 Depreciation Guide, Report 29, Volume 80, Chicago, Illinois.

[2] Energy Information Administration 1994, Electric Power Annual 1993, DOE/EIA-0348(93), U.S. Department of Energy, Washington, DC, December.

[3] Energy Information Administration 1996a, Input Data to the Annual Energy Outlook 1997, staff transmittal of electronic files, U.S. Department of Energy, Washington, DC, October.

[4] Energy Information Administration 1996b, Annual Energy Outlook 1997, DOE/EIA-0383(97), U.S. Department of Energy, Washington, DC, November.

[5] Hill, Lawrence J., 1995, A Primer on Incentive Regulation for Electric Utilities, ORNL/CON-422, Oak Ridge National Laboratory, Oak Ridge, Tennessee, October.

[6] Institute of Electrical and Electronics Engineers, 1995, An American National Standard: Guide for Distribution Transformer Loss Evaluation, Draft, IEEE, New York, April.

[7] Thomas, Steve, 1996, “Electric Reform in Great Britain: An Imperfect Model”,Public Utilities Fortnightly, 134(12), pp. 20-25, June 15.

[8] Pleat, George R., “Pricing and Profit Strategies for a Stand-Alone Electric Distribution Company”, Public Utilities Fortnightly, 135(2), pp. 20-24, January 15, 1996.

[9]. U.S. Advisory Commission on Intergovernmental Relations, 1992, Significant Features of Fiscal Federalism, Volume 1: Budget Processes and Tax Systems, Washington, DC, PB-92-198860, February.

[10] John L. Fetters, CEM, CLEP, “Transformer Efficiency , Energy & Power Management” 2002 BNP Media.

[11] Vijay Bhavaraju “Electronics in Energy Conversion”, Energy & Power Management oct.20, 2002 BNP Media.

[12] Deborah Penn “Forming Community-Owned Electric Utilities” Energy & Power Management 2002 BNP Media

[13] The Potential for Energy Efficiency in the State of Iowa, S. W. Hadley, 2001, ORNL/CON-481, Oak Ridge National Laboratory, Oak Ridge, TN, June

[14] Measurement Practices for Reliability and Power Quality, John D. Kueck, Brendan J. Kirby (Oak Ridge National Laboratory), June 2004, ORNL/TM-2004/91

Abbreviations/Acronyms

RF = The Peak Responsibility Factor

FCR = fixed charge rate

ISO = Independent System Operator

EC = Energy Cost

TOC = Total Operation Cost

TVA = Tennessee Valley Authority