The All-In Cost Comparison

Comparing behind-the-meter power to grid power is where most analyses go wrong, because they compare the wrong numbers. The right comparison is all-in cost structure against all-in cost structure, run over the way you actually operate, not a headline rate against a headline rate.

Begin with the grid bill and take it apart. A large industrial or commercial electricity bill is not one price. It is at least four components. Energy is the per-megawatt-hour charge for the electrons themselves. Capacity or demand is a charge for the peak power you draw, usually set by your single highest short-interval peak in the billing period, and it can be a third or more of the total. Transmission and distribution recover the cost of the wires that carry the power to you. And riders, taxes, and adjustments layer on top. The critical feature is the demand charge: it is set by your worst quarter-hour, so two sites using the same total energy can pay very different bills if one has a spiky load and the other is flat. Grid power's headline energy rate can look cheap while the delivered, all-in rate is far higher once demand and delivery are counted.

Now build the behind-the-meter cost stack. It has three main parts. Capital is the installed cost of the plant, converted to a cost per megawatt-hour by spreading it over the energy the plant will produce across its life, which is why running hours matter so much: the same capital divided by more megawatt-hours yields a lower per-unit cost. Fuel is the running cost, and for a gas plant at high duty cycle it dominates; it is set by your gas price, your heat rate or efficiency, and whether you capture the waste heat. Operations and maintenance covers service, parts, staffing, and the emissions controls the permit requires. Add those and you get a levelized cost, the honest per-megawatt-hour number to set beside the grid's all-in delivered rate.

Several levers move that comparison, and knowing them is the whole game. The first is capacity factor. A plant that runs near-continuously spreads its capital thin and competes hard; a plant throttled to a few hundred hours by an air permit carries crushing capital per unit and rarely beats the grid on energy alone. The second is heat recovery. If you capture exhaust heat through combined heat and power, you get useful energy from fuel you already burned, which can cut your effective cost per unit dramatically and is often the difference between winning and losing against the grid. The third is the demand charge you avoid. Even a behind-the-meter plant that does not beat the grid on energy can pay for itself by shaving your peak, because eliminating a large demand charge is worth real money regardless of the energy rate. The fourth is fuel spread: the gap between your delivered gas price and the local power price. Where gas is cheap and power is expensive, on-site gas wins comfortably; where that spread is thin, it does not.

This is why the answer is so site-specific. In a region with cheap gas, expensive congested power, a flat continuous load, and a use for waste heat, behind-the-meter gas can beat grid power on pure energy cost and win before you even count speed. In a region with cheap grid power, costly gas, a spiky load, and no heat sink, the grid wins on cost and self-generation only makes sense for reliability or speed. The same technology gives opposite answers in the two places, and anyone who quotes you a single national number for the cost of behind-the-meter power is selling something.

There is one more cost most comparisons omit: standby and backup charges. If you build behind the meter but stay connected for backup, the utility will often charge you a standby rate to keep capacity available for the hours your plant is down. That charge is real and belongs in the stack; a plant that ignores it will underestimate its true all-in cost. We treat that charge in detail in the next module, but flag it here so your comparison is honest.

The takeaway is a discipline, not a number. Build both cost stacks in full. On the grid side, use the delivered all-in rate including demand and delivery, not the headline energy price. On the behind-the-meter side, use a levelized cost that honestly loads capital, fuel, operations, and any standby charge, at your real running hours and with or without heat recovery. Compare those two, at your actual duty cycle, in your actual location. Only then do you know which side of the meter your next megawatt should live on, and only then have you earned the right to add the speed premium the next lesson describes.