A data centre rejects almost all the electricity it buys as low-grade heat. conventional electric chillers uses more electricity to move that heat outdoors. heat-driven absorption chillers uses the heat itself to drive the cooling cycle, which sounds like getting a service for free. The catch is thermodynamic: the temperature of the rejected heat sets an upper bound on how much cooling a unit of heat can produce, and that bound is usually well below what a motor-driven compressor achieves.
With conventional electric chillers, a compressor lifts heat from the chilled-water loop to the ambient loop using mechanical work; the coefficient of performance can reach five or six on a mild day. With heat-driven absorption chillers, a refrigerant is separated from an absorbent by heat rather than by a compressor, and the useful cooling per unit of driving heat is typically around 0.6 to 0.7 for a single-effect machine using hot water. Higher-temperature heat enables double-effect cycles with better ratios, but direct liquid cooling of chips rarely delivers water hot enough to use them.
The useful cooling per unit of driving heat is set by the rejection temperature, and a single-effect absorption stage typically delivers a coefficient of performance near 0.6 to 0.7 rather than the five or six of a compressor.
Part One: What is actually being compared
Before comparing heat-driven absorption chillers with conventional electric chillers, it helps to name the unit of account. In data-centre cooling load shifting the tempting unit is the headline specification, because it is printed on a datasheet and can be ranked. The unit that actually appears in a budget is the sum of three things: the cost paid before the first unit of value, the cost paid on each unit, and the cost of the occasions when the system behaves badly. Once those three are written on one line, most of the specification argument stops being interesting.
The cost structure diverges in a way that a per-unit comparison hides. With heat-driven absorption chillers, a meaningful part of the spend is committed before the first useful result; with conventional electric chillers, the spend arrives as a variable charge attached to each operation. A committed cost rewards volume and punishes a bad quarter, because it stays even when utilisation collapses. A variable cost is gentler operationally but removes the incentive to build spare capacity before it is needed.
Part Two: The arithmetic that decides it
The arithmetic is about what is being displaced, not what is being generated. If a 100 MW campus rejects heat at a temperature that supports a 0.65 COP absorption stage, and that stage offsets 15 percent of the mechanical cooling load, the electrical saving is roughly a few megawatts of compressor draw. Against that sit the pumping power for a much larger heat-rejection loop, the extra heat exchangers and the footprint of the absorption plant. On many sites the net PUE improvement is a few hundredths, not the step change the phrase waste-heat cooling suggests.
The mechanism behind the trade is about information rather than physics. heat-driven absorption chillers makes one part of the problem observable and therefore cheaper to plan around, while conventional electric chillers leaves it implicit and pays for it in variance. Whenever a design makes something measurable, somebody still has to build the instrument, keep it calibrated and act on its readings, and those three obligations are where a theoretical advantage is usually spent.
The second-order consequence is organisational rather than technical. Choosing heat-driven absorption chillers creates a dependency on a capability that must be staffed, monitored and kept current. Choosing conventional electric chillers keeps that capability outside the team and converts it into a vendor relationship. Both are real costs, and the option that looks cheaper in a spreadsheet is frequently the one that adds a standing meeting to the calendar.
Part Three: Where the simple model breaks
The real value appears at the grid boundary rather than inside the building. Compressor electricity demand peaks on the hottest afternoons, exactly when the grid is most expensive and most constrained. An absorption stage that is heat-driven flattens that peak without a battery, and the avoided demand charge or capacity payment can be worth more than the energy saved. But that benefit only materialises where the tariff rewards peak reduction, which is a commercial fact about a specific site, not a property of the equipment.
Offsetting roughly 15 percent of a 100 MW campus mechanical cooling load saves a few megawatts of compressor draw but adds a larger heat-rejection loop, more pumps and more evaporative water.
Part Four: The cost that arrives later
Water is the second bill. Absorption machines reject more heat per unit of cooling than a compressor cycle, which pushes cooling towers toward higher evaporation. At a site where water rights are contested, trading electricity for water is a poor bargain; at a site with abundant reclaimed water and a carbon-intensive grid it can be a good one. This is why two nominally identical campuses can reach opposite conclusions from the same vendor proposal.
Expect a split by scale rather than by industry. Large operators can amortise the fixed side of heat-driven absorption chillers and will move first; smaller ones keep conventional electric chillers because their load is spiky and their tolerance for operational complexity is low. The genuinely contested ground is the middle, where a managed service can make either option look like the other.
Part Five: What this changes outside the engineering team
The everyday consequence is about the neighbourhood. A campus that cuts its summer afternoon peak by a few megawatts reduces the chance that a local substation is upgraded, that a peaker plant runs, and that residential customers see a critical-peak price alert. Those are visible, mundane benefits, and they are often the reason a planning authority approves an expansion that would otherwise be rejected on load grounds.
Outside the engineering team the consequence is about who gets access. A system that needs less of the scarce resource can be replicated by organisations that could not have afforded the original, which changes procurement, staffing and the kinds of work that get scheduled. That is a quieter but more durable effect than any single performance number.
For a site evaluation, the practical question is the temperature of the available heat, the tariff's peak-demand component and the local water price. Those three numbers decide the outcome long before the chiller's rated efficiency does.
A market sizing sanity check
A plausible sizing: if heat-driven cooling and heat-reuse retrofits reach even a modest share of large campuses by 2032, the equipment, integration and controls segment plausibly sits in the $3B to $9B annual range. The lower bound counts only purpose-built absorption and adsorption plants attached to new facilities; the upper bound includes retrofits, heat-loop upgrades and the controls that arbitrage tariffs. The binding uncertainty is grid tariff design in the regions where new capacity is being built, not the thermodynamics.
All figures here are an author's synthesis of published ranges and a stated model, not a measurement of a shipped system. The point of the arithmetic is to locate the crossover and expose which assumption drives it. A vendor quote will differ, and the direction of that difference is usually more informative than the final number.
The comparison that matters
The comparison that matters is between compressor electricity avoided and the heat-loop, water and footprint costs added. Absorption wins on sites with a usable heat temperature, a peak-sensitive tariff and cheap water; electric chillers win where heat is too cool, water is scarce or the load profile is flat. The deciding variable sits outside the machine.
The honest caveat is that this ranking is sensitive to one assumption: that the workload keeps its present shape. Shift the mix toward the case where conventional electric chillers is strong and the result changes without any technology changing at all. That is why the durable output of an evaluation is not a verdict but the threshold at which the verdict would flip.
Decide on the available heat temperature, the peak-demand tariff and the local water price before comparing rated chiller efficiencies.
SOL: HCqPmPJM3NVtS6YrmVaTZYnuP8XZ2PFgQZKT94VK1LkS
Creating this takes real effort — thank you for the support.
Top comments (0)