Reciprocating Engines and Gas Turbines

The two combustion technologies contending for behind-the-fence duty are reciprocating gas engines and gas turbines. They burn the same fuel, natural gas, but they are mechanically and commercially different animals, and the differences map directly onto the four criteria we established: speed, firmness, cost, and emissions.

A reciprocating gas engine is, conceptually, a very large version of the piston engine in a car or a ship, running on natural gas and driving a generator. Individual units come in the single-digit to low-double-digit megawatt range, and the standard deployment pattern is to array many of them in parallel — a plant might be dozens of engines side by side, each in its own enclosure. This modularity is their signature advantage. Because each engine is a self-contained, factory-built unit, they can be manufactured, shipped, and installed quickly, and a site can be brought online in phases: energize the first bank of engines while later ones are still being installed. For a developer racing a compute deadline, that phased, modular character is worth a great deal. Reciprocating engines also excel at load-following — they can ramp up and down quickly and start fast, which matters when the load profile is spiky or when the engines are backing up an intermittent resource. Their electrical efficiency at the individual-unit level is strong, and critically, it holds up well in hot weather and at altitude, conditions that degrade turbine output.

Gas turbines work on a different principle: air is compressed, mixed with gas, combusted, and the hot expanding gas spins a turbine that drives a generator — the same basic machine as a jet engine, adapted for stationary power. Turbines come in a wide range, from smaller 'aeroderivative' units up to very large frame machines producing hundreds of megawatts each. Their great advantage appears at scale and especially in combined-cycle configuration, where the turbine's hot exhaust is routed through a heat-recovery steam generator to drive a second, steam turbine. A combined-cycle plant captures energy that a simple-cycle machine throws away, pushing overall electrical efficiency well above what reciprocating engines achieve. If your site is very large, your timeline permits, and you can site a steam bottoming cycle, a combined-cycle gas turbine plant is the efficiency champion — meaning less fuel burned, and less carbon emitted, per megawatt-hour.

But turbines carry offsetting burdens. A large combined-cycle plant is a substantial construction project with a longer lead time than a bank of engines, and the market for large turbines has itself become supply-constrained as demand surges — order books at the major manufacturers have stretched out, so 'just buy a turbine' can mean waiting years for delivery. Simple-cycle turbines, which forgo the steam bottoming cycle for speed, sacrifice efficiency to deploy faster, landing somewhere between engines and combined-cycle on every axis. Turbines also lose output in hot weather because hot air is less dense, a real consideration for data centers in warm climates.

On emissions, the picture is nuanced and matters enormously for permitting, which is the subject of Module 4. Both technologies emit carbon dioxide from burning gas, but the regulated criteria pollutants — especially nitrogen oxides, or NOx — behave differently. Reciprocating engines, because of how combustion happens inside a piston cylinder, tend to produce more NOx and other pollutants per unit of fuel than turbines do, and therefore lean harder on after-treatment systems such as selective catalytic reduction to meet limits. Turbines generally run cleaner on NOx at the source, particularly modern low-emission combustor designs. This is a genuine tension: the technology that deploys fastest, the reciprocating engine, often carries the heavier local-emissions burden, which can slow its permitting in a strict air district. That is Faraday's read on why you cannot evaluate the machines purely on megawatts and dollars — the emissions profile feeds straight back into the timeline through the permit.

So how does a developer actually choose? The honest answer is that it is site-specific, and the sophistication is in matching the machine to the situation. Need power fast, on a moderate footprint, in a place where you can meet emissions limits with after-treatment? Reciprocating engines are compelling. Building a very large campus where you can afford a longer construction window and want the best fuel efficiency and lowest carbon intensity? Combined-cycle turbines earn their keep. Facing a strict air district where NOx is the binding constraint? The cleaner-at-source turbine, or a fuel cell, may be the only path. In practice, large projects increasingly blend technologies — engines for fast first power and load-following, turbines for efficient baseload — because the criteria pull in different directions and no single machine optimizes all four at once. The lesson to carry forward is that these are not interchangeable commodities; each is a distinct bet, and reading which bet a developer placed tells you what they prioritized — and what constraint they were most afraid of.