Energy players can lose a lot more than their shirt if they fail to model transmission losses properly.

It's no secret that the delivery of electricity from the generator plant to the end user causes energy losses in the power system. Since 1960, losses typically have ranged from 8.5 percent to 10.5 percent of total generation annually. If the national average retail price for electricity is used, these losses have a value of $25 billion-a staggering sum that evaporates into thin air if not properly modeled, tracked, and accounted for. However, a close look at the information technology (IT) platforms of power marketers, off-the-shelf transaction management systems, and IT systems used by energy service providers shows that loss modeling in many of these product suites is very rudimentary, if not non-existent.

Each market has its own methodology for calculating losses. Market participants doing business in a given market must therefore have an intimate understanding of how different markets model losses, and in turn have an IT system capable of accounting for losses using the given market's methodology.

Most market participant IT systems reflect a rather casual regard for the importance of losses in the energy trading business. The financial implications of not properly accounting for losses can be significant.

The Cost of Losses

Energy losses in the transmission and distribution (T&D) systems must be replaced by additional generation resources. Two factors determine the cost of the system losses: the initial cost of energy used to generate what is eventually lost, and the cost of the added generation needed to replace that loss. The former is often referred to as the energy cost of losses, and the latter cost is often termed a demand cost. Both are expressed in $/kW. The demand cost is the marginal cost of additional capacity in both generation and transmission/distribution facilities needed to provide for the system losses.

The cost factors for power loss replacement are the same as those for the basic energy service-capital, fuel, and operations and maintenance components. Usually, transmission customers pay for losses on a system-wide or zonal-specific $/MWh basis, depending on the system variable operating costs. Traders conducting wheeling transactions can either pay for the losses, or they can supply extra power to make up for these losses. Similar to transmission losses, the energy lost in the distribution system must be supplied by the generators at the system variable operating costs.

In 1996, Oak Ridge National Laboratory developed estimates of ancillary service costs using data, assumptions, and analyses from twelve U.S. electric utilities. The estimates show aggregate costs of ancillary services ranging from $1.50 to $6.80/MWh, with an average of $4.15/MWh for the utilities sampled. The power loss replacement is the most expensive service, accounting for 30 percent of the total ancillary service costs. Thus, the energy losses in the T&D system had an average economic value of $1.25/MWh in 1996 dollars.

According to the Energy Information Administration, two-thirds of transmission losses occur in the more than 2.6 million miles of transmission and distribution lines that make up the national transmission system in the United States (155,000 of those miles being transmission lines of 230 kV and above). The remaining one-third of loss occurs in transformers. The losses in the transmission and distribution system as a whole account for about eight percent to 10 percent of the total net generation, with 55 percent occurring in the distribution system, and the remainder in the high voltage transmission system.

The difference between the net generation and sales is often used as an indicator of the losses (or efficiency) of the delivery system for the electric power: using this definition, the losses in the U.S. transmission and distribution system amount to about 379 billion kilowatt-hours annually, or 10 percent of the net generation.

The 8.5 percent to 10.5 percent loss figure includes the real physical losses in the transmission and distribution systems, as well as non-sampling errors and data collection frame differences. The actual physical losses associated with wires and equipment in the transmission and distribution systems are therefore somewhat smaller. Figures typically quoted for the physical losses in the transmission and distribution systems are 7.5 percent to eight percent of the total net generation from power plants.

System losses normally are classified into two categories: load losses and no-load losses. The no-load losses are incurred 24 hours a day, while the load losses vary with the time of day.¹

Strategies for Accounting for Wires Losses

To account for losses accurately, we will first look at how front office systems need to model and track losses during deal ticket entry and eTag generation. Then we will examine how back office systems need to account for and settle losses.

In terms of front office systems like those of energy traders, the two main systems that must model losses accurately are: a) the deal capture system-in other words, the physical, bilateral transaction management system; and b) any programmatic market communication and bidding system that communicates with central market entities such as Independent System Operators (ISOs) and Regional Transmission Organizations (RTOs). The front office systems are the first point of entry for a deal, transaction, or any market bid submitted to the central exchange. It is very important that the trader, market participant, or scheduling coordinator have all costs pertaining to losses available while making that transaction. This way there are no hidden loss costs that the organization later incurs during the monthly invoice process.

Also, monitor the tariff: call the ISO, Control Area, or OASIS operator and identify the sections within the relevant market rules and procedures document(s) that deal with transmission and distribution losses. Some control areas model losses as a static percentage that is applicable for the entire control area. Others run periodic AC power flows and come up with seasonal loss percentages. Still other central markets offer different loss percentages for different tie-lines. If that is the case, energy traders need to monitor Web sites of the central exchange to make sure they are working with the latest loss percentage.

Modeling the different varieties of losses in a trading operation's deal capture system also can be another effective approach. Such modeling permits the deal ticket entry screen to allow for inclusion of loss parameters based on different loss modeling calculations. For example, a pull-down menu might allow the trader to model losses using:

• A static loss percentage for the entire control area (this is the easiest method of handling losses during deal entry);
• Seasonal loss percentages;
• Energy payback schemes (also referred to as 7-day payback or 168-hour payback schemes); or
• A distribution factor matrix to compute different losses for each combination of points of receipt (POR) and points of delivery (POD).

In addition, some control areas require that traders provide for losses during the same hour that they schedule the energy, while others require traders to map the lost megawatts using a completely separate electronic tag. Others allow procurement of losses from third-party independent power producers and delivery of them to points in the grid other than the POR/POD listed on the main transaction. It is therefore very important to understand how the counter-party (in this case the ISO/Control Area/Transmission Provider) will require payment for losses. By mapping the business processes of interacting with each central market entity, it will be easy to have an IT platform automate most of that interaction.

Back Office Systems Accounting Techniques

Power losses are associated with all wholesale transmission service. The transmission provider is neither obligated to provide power losses, nor is the central market going to pay for the losses. The market participant ends up paying for these losses charges. It is especially important to ensure that back office systems track losses and account for them accurately, so there are no surprises with the monthly invoice from the control area. The ISO or RTO may have different schemes for recovering loss costs. Some ISOs recover losses through the marginal loss recovery mechanism, while others allocate losses based on a load weighted ratio (i.e., you pay for losses as a percentage of your serviced load as a ratio of the entire ISO system load), and yet others use concurrent energy payback schemes.

That is why there should be a feedback loop between back office invoice and front office loss parameters. The trader makes a trade based on interpretation of the loss charges in the transmission tariff document. That number may not always be the same as the total charge from the control area that makes its way into the month-end invoice. Loss percentages may be adjusted because of a more recent power flow run of the control area, or they could be adjusted based on payback charges, for example. Market participants need to know when there is a change in loss charges, so it is important to provide for the communication of these changes.

Some central markets allocate loss charges by using a load-ratio adjusted factor. In situations like these, risk management systems or transaction review systems have to deduct the loss costs from the identified bid-ask spread.

Perspective on Losses

Losses in the transmission and distribution system are inevitable, subject only to the laws of physics. However, these losses have significant economic value, and they require additional generation capacity to compensate for them. The losses, as defined earlier, amounted to 379 billion kWh in 1999. This is equivalent to 43,000 MW of generating capacity, or five percent of the presently installed generation capacity in the United States.

Even small improvements to the efficiency of the transmission and distribution system will have an immediate impact on energy markets. With significant increases, however, energy traders will still be well advised to understand how losses are modeled in the markets in which they do business. The dollars associated with losses on any single deal might not raise many eyebrows, but the fact that these negative financial implications are avoidable is reason enough to incorporate accurate loss modeling into the trader's IT systems. As competitive markets evolve, margins get thinner and arbitrage opportunities grow scarcer, the ability to account for losses precisely will become increasingly important to the trader's bottom line.

1. It is also important to distinguish between the losses that occur physically within the distribution system and the losses that are associated with inaccurate metering of energy, but this article will not focus on this distinction.

Wholesale electric power markets use a variety of methods to model transmission losses. Following is a summary of the major U.S. markets' methodologies. Market participants doing business in a given market should have the capacity in their IT systems to shadow the calculations performed by the ISO/RTO, to have a precise understanding of what losses are costing them.

California ISO

The California ISO uses Tie Meter Multipliers (TMM) data, along with the final hourly schedules to compute the dynamic loss factors that are applied for a particular bilateral transaction or import transaction into California. Once the TMM is identified, then the market participants, known as Scheduling Coordinators, are charged the incremental cost of energy at that location to cover the cost of losses. This means if the TMM is two percent and the import transaction into California is 100 MW, the cost of providing the two percent losses will be charged at the marginal cost of energy for two MW at the delivery location. The TMM for each hour in each location are posted on Cal ISO's OASIS site.

West Connect (formerly Desert STAR)

West Connect plans to handle losses with concurrent energy payback. Concurrent energy payback is a scheme (more predominant in the Eastern Interconnection than the West) whereby the customer of the control area is required to provide for the loss megawatts at the same time the transaction is scheduled. The customer has the option of: a) providing for the loss megawatts themselves; b) buying the loss megawatts from a third party power marketer; or c) choosing to buy the loss megawatts from the transmission provider itself. However, financial compensation for losses is not allowed.

ERCOT

T&D losses are the responsibility of the Qualified Scheduling Entity (QSE) representing the competitive retailer's load. The QSE has the responsibility to schedule the necessary megawatt amount of energy to cover the retailer's load plus applicable T&D losses. The ERCOT ISO forecasts the ERCOT-wide transmission loss factors, expressed as a percentage of load for each settlement interval of the operating day. Distribution loss factors by settlement interval are determined by each Distribution Service Provider (DSP), which in turn submits them to ERCOT. The settlement process uses the forecast loss factors and deemed actual loss factors when adjusting aggregated load for losses to determine total QSE load obligations. Seasonal transmission loss factors are derived from annually updated ERCOT on-peak and off-peak load flow base case analysis conducted by ERCOT. The ERCOT ISO makes the transmission loss factors available on their Web site (http://www.ercot.com).

ISO New England

Losses on the NEPOOL transmission system, especially the Pool Transmission Facilities (PTF), are determined for each market participant's and non-participant's generation obligations for each hour. The obligations are adjusted automatically for their share of losses (net losses after through losses and out losses have been deducted). The responsibility for losses is included as an adjustment in the settlement obligations via additions to the generation or load obligation.

New York ISO

NYISO addresses transmission losses by computing the marginal (or incremental) effect of real power transmission losses for the day-ahead market (aka BME), and real-time operations. Balance Market Evaluation refers to a process that NYISO runs on an hourly basis to tune their unit schedules. There is no market for this like there is for day-ahead and real-time, but the ISO does consider losses with regard to all three. NYISO calculates a penalty factor, which is defined as the increase in generator output at a certain network bus that is required to supply an increase in load, taking into account network transmission losses. Generator energy bids are multiplied by penalty factors to account for incremental transmission losses in the dispatch process. NYISO also uses marginal loss calculation algorithms to compute losses associated with external transactions flowing through the New York Control Area. The locational marginal pricing (LMP) algorithm identifies the energy cost, loss cost, and congestion cost components separately.

PJM

Point-to-point customer losses are financially quantified and charged at the hourly PJM load-weighted-average LMP based on average loss factors of three percent for on-peak and 2.5 percent for off-peak periods. The loss revenues from point-to-point customers are allocated as credits to network customers based on hourly load ratio shares. Network customers, on the other hand, have their losses implicitly included in their load quantities except for 500kV losses, which are metered and allocated to fully-metered electric distribution companies based on hourly load ratio shares.

WECC Region (formerly WSCC)

Presently within WECC, the most common practice is to require energy payback as a means of dealing with losses. Energy payback is a scheme in which the counterparty does not explicitly pay for transmission losses during the bilateral transaction or during the scheduling hour, but wheels back the amount of MW dissipated in losses within a pre-determined length of time or during an off-peak period. If the payback is within one week, it is referred to as the "168-hour energy payback scheme." For some control areas, energy payback is a necessary scheme because these areas rely on the lost energy to serve their native load needs. This seven-day barter of energy is more practical than requiring the control area to purchase lost energy from the spot market and charge the customer for it.

Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.