In my previous Editor's Comments, I suggested that there is no such thing as having too many options—a belief that has saved my bacon numerous times in the past. This fetish has spawned a lifelong quest to develop an anthology of quick and dirty rules of thumb that help me better understand and deal with a variety of situations, some of which might seem pretty mundane but are useful nonetheless.

For instance—in case you are interested in such things—tire hydroplaning speed is a function of tire pressure. At water depths in excess of your tread depth, it will occur at roughly 10 times the square root of your tire pressure. Thus, it's a good idea to visit your tires from time to time, checking both their condition and inflation, and then when you run into a deluge, you can say, "Aha! I'm good for 60." Well, give yourself a couple of knots for the wife and kiddies, but—simplistic as this might seem, compared to the rather lengthy series of equations needed to reach a surprisingly similar conclusion—it has the indisputable advantage of being hauled out instantaneously and put to real-world use.

Now we come to the point of this column, where I asked several people from the distributed-energy industry to help me develop a suite of benchmarks to aid my understanding of relevant issues. What I found is that there are no simple answers to anything having to do with energy, yet I was able to get around to at least a few approximations in response to several questions I posed. To some, these questions and answers will be old hat, and to you who view them as such, I hereby offer a standing invitation to help me pose better questions for future columns. Others of you may disagree with the following for a variety of reasons—ranging from "newer equipment is better than that" to "that's nice in theory, but I've never seen it work that way in practice"—and from you I want feedback, no matter how blistering. My e-mail address is editor@forester.net, and, barring catastrophe, the sun never sets on our server.

For years, people assumed 33% to be the average efficiency for central generation with no heat recovery, representing a span from less than 20% for peaker systems up to 55% for combined-cycle gas turbine units.

Since most new generation is combined cycle, one would expect that figure to move rapidly toward the higher end of the efficiency range.

While in theory all of the same generation technologies are applicable, others exist (solar, fuel cell, etc.) that are inherently more appropriate in smaller sizes. In practice, local-level installations can be more efficient than many grid-based systems principally because of increased heat-recovery opportunities.

While heat recovery at central plants is possible, the cost for conducting steam to a thermal customer more often than not renders the practice uneconomical. For most onsite projects, however, not only does the potential for heat recovery exist, it is in many cases an economic necessity. A simple calculation should suffice to prove the point:

At present, the average retail electric rates are in the neighborhood of $0.67/kWh, and average natural-gas prices are roughly $6.00/MMBtu. Thus, if you want to convert retail gas into retail electricity, you must have a minimum efficiency of 11,666 Btu per kWh, or about 30%: (.067 ÷ 6x 1,000,000) ÷ 3,413 (3,413 is the number of British thermal units per kilowatt-hour).

In practice of course, you will need to consider something in the range of $0.005–$0.20/kWh to cover operations and maintenance costs—plus perhaps another $0.02/kWh for capital recovery, contingency, and profit. Considering these aggregated costs, heat recovery is a valuable component of an increasing number of distributed-generation projects.

In many cases, onsite generators are sized to "opportunity" fuel supplies (e.g., biomass, waste heat, solar power, blast furnace gas, etc.) and while there certainly are costs associated with their exploitation, often their use does not entail marginal fossil fuel–consumption costs.

What overall efficiency increase can be expected where heat recovery is involved? It depends on the amount of heat recovered, but the simple calculation of the impact can be made based on the following series of equations:

1. Efficiency (%) = 3,413 ÷ heat rate (in Btu/kWh)
2.Waste heat = heat rate x [1 - efficiency (%)]
3. Recoverable heat = waste heat x heat recovery (%)
4. Overall efficiency = [heat rate x efficiency (%) + recoverable heat] ÷ heat rate

Assume for simplicity's sake that we have a gas turbine with a heat rate of 10,000 Btu/kWh. Since there are 3,413 Btu/kWh, from the above equation this means that the gas turbine has an overall efficiency of 34%. Thus, in this scenario 6,587 Btu of the energy value of the fuel is being sent out of the stack as waste heat.

It's not economically possible to recover 100% of the waste heat, but systems exist that allow you to get about 75%, so you could boost the overall efficiency by recovering 4,940 Btu/kWh, or 49% (66% x 75%), of the input fuel. Whether or not you actually recover this much of the fuel depends on local thermal needs, but this at least establishes a practical upper limit for recovery.

Since this heat recovery displaces heat that you otherwise would have needed to power an onsite boiler, you can assume that this would displace another combustion process with an efficiency of about 90%, since 0.9 unit of recovered heat displaces 1.0 unit of purchased fuel.

A summary of the economic—and environmental—impact of this 10,000 Btu/kWh gas turbine with 80% heat recovery follows:

Central plants typically show only the marginal cost of generation on the assumption that they will continue to operate as long as they are recovering their fuel costs. Distributed-generation plants often build capital recovery into their cost structure. Both are valid calculations but answer different questions and should never be compared directly.
One of the problems in achieving an apples-to-apples comparison is demonstrated effectively in the analysis of wind turbine economics. Many experts claim that wind-turbine electricity runs at $0.04–$0.06/kWh, which leaves the impression that wind isn't quite competitive yet. But if you accept the premise that the marginal cost is zero and thus every marginal kilowatt-hour from a wind farm displaces a more expensive marginal kilowatt-hour somewhere else, the valid comparison would seem to be the speed of capital recovery.
For other technologies, the efficiency basically runs the same span as it does for central generation—some low-end installations at 20–30% without heat recovery, lots of technologies in the 30–40% range, and an increasing number in the 40–60% range (e.g., combined-cycle gas turbines and some fuel cells)—all of which can be boosted with heat recovery.

A case can be made that even if centralized and onsite generation were economically equivalent, onsite generation would still come out ahead since it doesn't require additional investment in wires. According to FERC, the average United States transmission-and-distribution investment costs $1,300 per delivered kilowatt. This, in a nutshell, explains the difference between wholesale power prices ($0.02–$0.04/kWh) and retail rates ($0.06–$0.20/kWh). In addition to capital are line losses that average 9.5% in the US but can reach as high as 20% during peak periods. Thus, a kilowatt-hour with a marginal cost of $0.03 at a central plant can have a marginal cost as high as $0.0375 [3 ¸ (1 - 0.2)] at the end of the wire during peak periods.

So what have we learned from all of this? Probably that while no simple answers exist in comparing centralized- and onsite-generation schemes, there still are some thumbnail observations we can make and that the same caveats apply to comparisons between various onsite-generation options.

Send John an email

DE - Jan/Feb 2004