Coal, Gas

Turn Emergency Generators Into Dollars

Issue 10 and Volume 101.

Turn Emergency Generators Into Dollars


Thomas P. Sheahen and Gilbert R. Stegen,

Science Applications International Corp.

The Concept of Distributed, Dispatchable Power generation is essentially the reverse of interruptible

service.1 It can be understood by regarding both power and money as vectors: when the direction of the power flow switches, so does the direction of the money flow. At a signal given by the utility, a factory activates its emergency generating system and briefly becomes an independent power producer (IPP), feeding power into a local region of the grid. Upon receipt of another signal, it retires from that role. It may, however, continue to generate power for its own use.2


Many factories, office buildings and hospitals have standby generators, which are normally used only when there are power outages.3 Typically, they are sized smaller than the user`s full-load requirements. During an outage the user shuts down non-essential uses and connects the generator to the downsized load. When the grid can again supply full power, the standby generator is disconnected and the full load is reconnected to the grid.

Because the power seldom fails, these standby generators almost never go into action. Therefore, a significant amount of capital is tied up in idle equipment. The underlying economic strength of this proposed concept is in converting that idle equipment into increased capacity. This provides a greater safety margin for both the utility and the factory or other facility.

From the utility`s point of view, during times of high power usage the grid may not be able to supply sufficient power to maintain the desired load. In this situation, power generated at a customer`s site reduces the apparent load. The value of this power becomes very high when the grid is at or near full capacity. During this period, any significant load perturbations could cause grid collapse. If the standby generator is connected seamlessly to the load, thereby reducing the apparent load, the grid demand will be reduced and a safe operating margin maintained.


It is not trivial to interconnect a series of auxiliary generators in this way; controlling the entire system and dispatching selected generators requires sophisticated system engineering. The control and dispatch enterprise could be run by an intermediate energy-management service provider, or by an office within the utility itself. One such service provider is the team of Science Applications International Corp. (SAIC) and AuBeta Technology Corp., a company that has developed a proprietary dispatchable standby generation system controller. Absent this controller, it would be impossible to keep a collection of distributed generators under control.

There are three elements of con trol involved in a system of this type:

First, the standby equipment must be synchronized with the existing grid when it comes on. This is in contrast to the customary operation of standby generators, which are connected break-before-make, and hence do not require grid synchronization. Figure 1 illustrates the concept. Accordingly, the dispatching system located at a remote site must be able to detect and control that synchronization prior to locking into the grid.

Second, the dispatching system must choose generators from one or more participating facilities to send electricity to where the need is greatest. This requires some sophis ticated knowledge of the grid, including the typical loads being carried by various distribution buses. It is important that increasing capacity in one sector during a time of high demand does not overload or otherwise upset some neighboring sector.

Third, the standby generation equipment must work upon demand. To ensure this, the energy management service provider takes over complete responsibility for the maintenance of the equipment at each location. Regular maintenance visits, coupled with continuous monitoring and testing from the remote dispatch center, combine to assure reliability of the equipment.

In practice, a system is set up as follows: The energy management service provider secures the rights to control generators. In exchange for those rights, the generator owners receive financial compensation, usually proportional to the size of the generator and the expected output. Next, the provider negotiates a contract with the local utility to provide peak power capacity in specific geographical regions. This is a very important element: the utility will buy peak power inserted at a certain location. The utility pays the energy manager for this specialized extra capacity in proportion to the total capacity.

The energy manager designs, installs and operates the special equipment needed to connect the generators to the grid, and the additional equipment needed to remotely control and monitor each generator and its power production.

Figure 2 shows how both the electrical and the money paths are linked in this system. The money reaching the generator owner amounts to a long-term lease on the possibility of using his equipment, through which the utility can meet its other needs.


From the factory`s point of view, this concept means “found money.” The cost of the standby generator is sunk capital, long since justified by emergency considerations. A third party`s offer of money just for the capacity–the right to turn on its generator–is an offer that`s difficult to refuse. The factory further benefits from the reduced cost of maintaining its standby equipment. Moreover, at such time as the standby generator does turn on, the entire fuel cost is borne by others. And in a true outage or emergency, the standby equipment reverts to the factory`s own control.

The power generated falls in the category of peak-shaving, but that is not sufficient motivation to interest a utility in participating. The availability of wholesale wheeling suggests that there will always be cheaper power available than power produced from standby generators in factories. But it is important to remember that the grid is subject to transmission/distribution (T/D) capacity limits. There are many urban regions where the utility cannot deliver additional power from external sources without overloading the system. Therefore, utilities would like to have peak power “home grown,” i.e., both generated and used just a few miles apart. If that can be done, the utility avoids the cost of upgrading the T/D system, a modification both expensive and fraught with opposition in many urban settings. Therefore, the utility may find it economically advantageous to acquire additional capacity that is located within the highest-demand sectors of its grid.

Cost considerations

Determining the cost of added capacity is easier than determining the value of added capacity. At times when the grid is in danger of collapsing, the value is much higher than in times of general excess capacity. Conversely, if the grid has capacity but the system is short of power and additional power can be bought from other sources, then the value is higher but not high enough to motivate a system such as this.

There are additional factors that affect both cost and value of power. For example, guaranteed uninterruptible power carries a premium. In the case at hand, the value of additional capacity will depend on which part of the power delivery system is operating at or near full load. When the entire T/D system is operating well below full capacity, the incremental cost is only the cost of power production. When the distribution system is operating at or near capacity, the avoided incremental cost for the next amount of capacity may be very large.

In general, at every utility there will be some number representing the cost of additional generating capacity. For simplicity, assume $1,000/kW, but with the important proviso that the cost of T/D expansion can easily double this number. Tacking on further environmental costs, a new gas turbine facility might exceed $2,000/kW. By comparison, the necessary instrumentation and control of an already-installed diesel generator costs from $100 to $250/kW. This constitutes a substantial incentive, and herein lies the opportunity for this dispatchable, distributed generation system to be economically viable.

Of course it is necessary to “annualize” the cost of new capacity. An investment of $2,000 corresponds to a foregone annuity of well over $100/yr, so the annual value of having an extra kW capacity likewise exceeds $100/yr. If the energy management service company charges the utility only half that, it is a good deal for the utility.

Next, some payment must be made to the standby generator`s owner. Here we hypothesize a yearly payment to the generator owner of $4.80/kW, plus having the standby generator maintained. Receiving $4.80/kW every year for just the right to turn on his emergency generator occasionally is a very good deal for the facility`s owner.

At first it may seem extreme to pay the factory only one-tenth what the utility is charged. However, it must be remembered that the monitoring and control equipment has a substantial cost that the service provider needs to recover over a period of time. $125/kW is used as a mid-range number for the cost of instrumentation and controls on the standby generators. Some sample cash flow calculations, to illustrate the economics, are shown in Table 1.

In this scenario, the service provider`s annual gross income is about one-third the cost of the monitoring equipment. This works out to a simple payback period of roughly three years, or an internal rate of return (IRR) of 27 percent. When maintenance costs are included (perhaps $2/kW annually), the effective payment to the generator owner rises to nearly $7/kW annually, the payback period to the service provider stretches out, and the IRR drops.

The point of these illustrative calculations is to show that the energy management service provider also experiences an attractive deal, without which the entire plan would never get off the ground. Obviously, the provider is motivated to reduce capital cost (the monitoring and control equipment) and thereby enhance profitability.


What could possibly go wrong?

The first thing that comes to mind is the conservatism of utilities. Nothing can be hung on their systems until it is proven to be trouble-free. Unless the utility has complete confidence in the energy management service provider, this concept will never be implemented. Very careful engineering tests must support each stage of this system.

A second significant obstacle is that this system is not needed everywhere. The great majority of peak-capacity needs can be met with power wheeled from far away. For many utilities, peaking capacity might be several GW, perhaps through a pumped-hydro facility. Clearly, a few MW in a system of this type is tiny. Furthermore, the cost of combustion turbines is dropping rapidly. Therefore, only in special geographical locations where the T/D system is strained (mostly dense urban regions) will utility managers consider the desirability of having the peak power generated within that local region of their service territory.

By far the most severe obstacles are institutional rather than technical. IPPs have already experienced great difficulty in obtaining air emissions permits. In general, standby generators have customarily been exempt from such permitting requirements. When this concept was first suggested in a metropolitan region of northern Virginia, the local regulatory board ruled that any unpermitted standby generators would have to be permitted. Immediately, this drove many potential participants away. In order to retain an exemption from permit requirements, it would be necessary to convince the environmental-quality regulators with authority over an urban region that the only time these standby generators would go on would be in a genuine emergency, as the last measure just short of a blackout.


To date, proof-of-concept and a number of engineering tests have been carried out by AuBeta Technology Corp. in cooperation with Puget Sound Energy in Washington State. Control software was tested and a small (5 kW) generator was successfully controlled and logged.

The next step was to test the controller system with a large (300 kW) generator. Cummins Northwest in Renton, Wash., has installed Onan parallel switching equipment designed to seamlessly connect the 300 kW generator to its internal load. This means the generator will be synchronized to the grid and then connected. Equipment is installed to guard against ground faults and other potential failure modes, such as unintentionally providing power to the grid. Typically, the generator will be powered up to carry the entire load from the grid before disconnecting. Power-quality monitoring equipment was installed to verify that the system was not perturbing either the load or the grid. This system is presently fully functional.

The next step is to instrument several generation facilities at different locations. The purpose of this is to have an ensemble of generators all working together. First, there will be a test of the signaling and control algorithms, without actually turning generators on. Next will come production of power kept entirely within the site (although with full synchronization), and only after that, electricity will be inserted into the grid. This step-by-step approach will increase confidence that the system will have no adverse effects on the grid.

Once this point is reached, questions of regulations and licensing will need to be settled. After success is demonstrated on a small scale, many factories may wish to participate. SAIC anticipates negotiating contracts with various generator owners and utilities to provide peak power. By that time, the need for locally generated power will presumably be even greater than it is now. S

Authors–Thomas P. Sheahen is a senior scientist in the Energy Systems Group Science Applications International Corp., specializing in energy efficiency of factories. He holds a doctorate degree in physics and is a registered Professional Engineer in Maryland.

Gilbert R. Stegen is an SAIC corporate vice-president responsible for developing new information technology applications in the power industry.


1 Lamarre, L., “The Vision of Distributed Generation,” EPRI Journal, April/May 1993, p 6.

2 Dellorto, G.E., “Co-Generation Increasing Reliability, Reducing Costs,” Power Quality Assurance, March/April 1995, p 48.

3 Middleton, J.R., “Emergency Power, More Than a Power Source,” Power Quality Assurance, May/June 1995, p 15.

Click here to enlarge image

Click here to enlarge image

Click here to enlarge image

Click here to enlarge image

Typical industrial generator that could be used as a distributed generation resource: 95 kW Waukesha generator at an oil field facility. Photo courtesy of Power Strategies.