Cogeneration (CHP) Savings Estimator
The Economics of Cogeneration
Cogeneration pays in two ways at once, and understanding both is the key to reading your estimate.
The first is the electricity you no longer buy. Whether that saves money depends on the gap between what you pay for power and what it costs you in fuel to generate it yourself.
The second is the heat you no longer buy fuel to make. Your CHP unit produces that heat as a by-product of generating the electricity. Every therm of recovered heat you can actually use is a therm your boiler does not burn.
The second stream is what separates on-site cogeneration from a power plant, and it is why a project can work at a site where generating electricity alone would not.
Spark Spread — Stated Correctly
The spark spread is the margin between the value of electricity and the cost of the fuel used to make it.
The EIA defines it as:
> Spark spread ($/MWh) = power price ($/MWh) − [natural gas price ($/MMBtu) × heat rate (MMBtu/MWh)]
The heat rate is the efficiency term: the fuel energy needed per unit of electricity. Because a more efficient machine has a lower heat rate, higher efficiency produces a wider spread. Fuel cost is divided by efficiency, not multiplied by it — a distinction worth checking on any calculator you use, since getting it backwards makes efficient equipment look worse than it is.
Two things matter when applying this to a building rather than a power station:
Your benchmark is a retail rate, not a wholesale one
You are not competing with wholesale power prices. You are avoiding the delivered rate on your bill, which is considerably higher — one reason on-site CHP can work in markets where merchant generation would not.
A plain spark spread ignores your heat
It measures electricity against fuel and stops there. It gives no credit for recovered heat, which for a well-matched building is a large part of the return. A spark spread that looks marginal can still support a strong project once the thermal side is counted. Our estimator counts both.
Note also that the spark spread is a gross margin. It excludes operations and maintenance, fuel delivery charges, taxes, and fixed costs. It is a screening signal, not a decision rule, and no single threshold value substitutes for pricing a specific site.
How Our Estimate Is Produced
The estimator takes your facility profile and energy costs and works through the following:
- Your monthly bill and rate give an approximate annual consumption and average electrical load.
- A unit is sized against that load, then matched to the nearest sizes we actually build.
- Fuel input, electrical output, and recovered heat are calculated from equipment performance at that size.
- Recovered heat is adjusted by how much a building of your type can genuinely use while the unit runs — the single assumption that most affects the result. The estimator states the percentage and the equivalent full-load thermal hours it used alongside every result.
- Avoided electricity and avoided boiler fuel are netted against fuel cost and maintenance.
Two details behind that worth stating, because they materially affect the numbers:
Efficiencies are stated on a higher heating value basis
Equipment datasheets frequently quote lower heating value, which produces efficiency figures roughly ten percent higher for the same machine. Mixing the two — LHV equipment efficiency against HHV fuel pricing — overstates savings substantially, and does so most on smaller units. We convert everything to HHV so the efficiency and the fuel price are on the same footing.
Electrical efficiency rises with unit size
A small unit converts less of its fuel to electricity than a large one, though it recovers proportionally more heat. Total fuel utilisation stays high across the range; the split between electricity and heat shifts. The estimator applies a size-appropriate figure rather than one blanket number.
Why We Show Ranges Instead of Exact Figures
Every result appears as a range — size, savings, payback, and CO₂. That is deliberate.
A screening estimate built from a monthly bill, a facility type, and an operating pattern cannot support a precise number, and presenting one would imply a confidence that does not exist at this stage. The range reflects reality: final figures depend on interval electrical data, actual simultaneous thermal load, your specific tariff, available incentives, and the equipment ultimately selected.
Treat the range as a screening signal that tells you whether a detailed analysis is worth commissioning. That is the decision this tool is built to inform.
What the Estimate Excludes
Stated plainly, because these are the items most likely to move the final number:
- Demand, standby and non-bypassable charges. The estimate works from a blended electricity-cost proxy derived from your bill and rate, so it does not separately model how CHP changes these. They need tariff analysis, and standby charges in particular can work against a project.
- Incentives, grants, and utility programs. Not included. These can be significant and vary by state and utility.
- Standby and interconnection tariffs. Reviewed separately; they can work against a project.
- The composition of your bill. The calculation assumes your bill is mostly usage charges. A bill with large fixed, demand, or rider components will behave differently.
The Environmental Estimate
Where a calculation is possible, we show a CO₂ range using the method behind the EPA's CHP Energy and Emissions Savings Calculator:
> avoided grid emissions + avoided boiler emissions − CHP fuel emissions = net reduction
Avoided grid emissions use eGRID subregion factors rather than an average grid mix, because what CHP displaces is fossil generation, not the nuclear and renewable share. Which factor applies depends on how the unit runs: above 6,500 operating hours a year it behaves as baseload and displaces the fossil fleet, so the fossil-fuel output rate is used; below that it is more likely displacing marginal generation, so the non-baseload rate applies. A transmission and distribution loss allowance for the relevant interconnection is added, since electricity generated on site never travels.
Two limits are worth stating. Some states span more than one eGRID subregion; where the split is genuine we show no carbon figure rather than guess, and where one subregion clearly dominates we say the assignment should be confirmed against the facility's ZIP code and utility. And the factors are reviewed annually — EPA republishes eGRID each year.
We report unfavourable environmental results too. On a relatively low-carbon regional grid — one with substantial hydro or nuclear generation — natural-gas cogeneration may deliver financial and resilience benefits without a material carbon reduction, and the tool says so rather than showing a number that flatters the technology. A carbon claim is only worth making if it survives the cases where it does not hold.
These figures are preliminary screening estimates. They are not intended for regulatory reporting, emissions inventories, or formal sustainability disclosures.
Frequently Asked Questions
It is a screening estimate, not an engineering analysis. It is built to answer one question: is a detailed review worth your time? The most valuable thing you can do to improve it is enter your actual rates from a recent bill instead of the regional averages.
