Adsorption Cooling — Performance, Working Pairs & Numbers

Adsorption Cooling — Performance, Working Pairs & Numbers

The quantitative deep-dive behind Adsorption Cooling, mined from the full-text PDFs in the adsorption set (47 of 113 items had local full text): the page to reach for when moving from “adsorption is interesting” toward sizing a real machine — what COP and cooling power to expect, from which working pair, at what driving temperature, and what it costs.

The quantitative deep-dive behind Adsorption Cooling, mined from the full-text PDFs in the hothothot/Joule Heist adsorption set (47 of the 113 items had local full text). This is the page to reach for when moving from “adsorption is interesting” toward sizing a real machine: what COP and cooling power to expect, from which working pair, at what driving temperature, and what it costs.

Two metrics run through everything:

• COP (coefficient of performance) = cooling delivered ÷ driving heat supplied. For thermally driven machines this is a thermal COP, not the electrical COP of a compressor — see the fairness note below.
• SCP (specific cooling power, W·kg⁻¹ of adsorbent) = how much cold per kg of sorbent per cycle. SCP sets the machine’s size and cost; COP sets its running efficiency.

The benchmark operating point

A recurring real-world test point across the literature — hot water 80 °C, cooling water 30 °C, chilled water ~14 °C — yields, for silica-gel/water:

• COP ≈ 0.45, SCP ≈ 176 W·kg⁻¹, ~4.3 kW from a lab chiller (Dynamic and Economic Investigation…; corroborated in Studies on the feasibility… and Adsorption Cooler Design…).

That is the honest center of gravity: COP in the 0.3–0.6 band, SCP on the order of 50–200 W·kg⁻¹ for conventional silica gel. Everything below is about where you land in (or beyond) that band.

Working pairs — the real trade

• Silica gel / water — the workhorse. Cheap, benign, low-temp drive (60–95 °C). COP 0.3–0.6, SCP ~50–200 W·kg⁻¹. Limited uptake → bulky. The default unless a reason to move.
• Zeolite / water — higher uptake and SCP (mobile double-bed hit 600 W·kg⁻¹) and stability, but classic zeolites need >150 °C regeneration — a real drawback that disqualifies low-grade waste heat.
• EMM-8 (zeolite-like aluminophosphate) — the standout in the set: COP 0.85 at a 63–65 °C driving temperature, water uptake 0.28 g·g⁻¹ at P/P₀=0.2, fast kinetics, hydrothermally stable. Drives 5–15 °C lower than reference materials for the same performance — i.e. it brings high COP into the waste-heat / low-temp-solar band where silica gel is weak. (Ultralow-temperature-driven water-based sorption refrigeration…)
• MOFs (UiO-66(Zr), CPO-27Ni) — very high SCP (up to 840 W·kg⁻¹) and dual cooling+desalination, but cost/stability/scale-up are open. The research frontier.
• Activated carbon / CO₂ (or NH₃, R-134a) — for sub-zero evaporator temps and pressurized compact systems where water (freeze, vacuum) is unwanted. See CO₂ (R744).

Selection is not obvious — Selection of a favorable zeolite for solar adsorption cooling makes the point that the “best” sorbent depends on the exact drive/sink/chilled temperatures; isotherm shape relative to the operating window matters more than headline uptake.

Squeezing more out: bed & heat-transfer enhancements

Adsorbents have poor intrinsic heat/mass transfer, so much of the engineering gain comes from the bed, not the material. Measured improvements from the set (Progress in design of adsorption refrigeration systems. Evaporators; others):

• Evaporator tube turbulators: COP +10.5% / SCP +9% (twisted tape); +41% / +47% (Z-type turbulators).
• Coated vs. uncoated tubes: COP +20%, SCP +47.6%.
• Conductivity additives: aluminum foams, 3-D graphene binder for zeolite 13X, metal additives, adhesive binders — all target the same heat/mass-transfer bottleneck.
• Cycle-level: heat & mass recovery between beds, and adaptive half-cycle timing, lift system COP (e.g. system COP/SCP 0.63 / 337 kJ·kg⁻¹ with recovery).

The COP-fairness argument (why 0.5 isn’t as bad as it looks)

A thermal COP of ~0.5 looks poor next to a compressor’s electrical COP of 3–5 — but that compares different energy currencies. Adsorption Cooler Design… makes the correction explicit: account for the thermal-to-electric conversion efficiency of grid power (η ≈ 40%), and vapor-compression’s primary-energy COP drops far below its nameplate — “which means AC systems’ COP is not a real disadvantage.” When the driving heat is waste or solar (primary-energy cost ≈ 0), the comparison tilts further. The strategic read stands: adsorption wins on free heat, loses on first cost and footprint.

Hot-climate derating (read this for San Antonio)

Performance is strongly sensitive to heat-rejection (condenser/cooling-water) temperature. The same machine that gives COP 0.50 / SCP 44 at a 30 °C condenser falls to COP 0.31 / SCP 26 at a 42 °C condenser — a ~40% capacity loss on the hottest afternoons, exactly when cooling demand peaks. Implication: a good heat-rejection sink is not optional in San Antonio — pair the chiller with geothermal ground-loop or a well-sized cooling tower, and oversize for the design-day condenser temperature, not the average.

Economics

• Solar-driven (residential): optimal collector ~38 m², feasible at collector unit cost < $700/m², life-cycle savings ~$3,500, payback ~11 years (An Economic Investigation…; Dynamic and Economic…).
• Jordan case: operating savings ~$5,000/yr, up to $20–30k/yr in a larger installation (…economic benefits for Jordan).
• First cost is the barrier — low SCP → large adsorbent mass and heat-exchanger area → high $/kW. This is why DOE/ARPA-E programs and the cost of manufacturing adsorption chillers work target SCP and bed cost directly. Free driving heat is what makes the payback close.

Cogeneration: cooling + fresh water

A major bonus thread — adsorption desalination + cooling (ADC): the same cycle yields chilled water and distilled water off the condenser. Reported specific daily water production (SDWP): silica gel 4 kg/kg, UiO-66 MOF 24 → 40 kg/kg, CPO-27Ni 6.8 kg/kg potable while also giving SCP 200 W·kg⁻¹ (Pore-size engineered nanoporous silica…; Adsorption Desalination and Cooling Systems…). Two utilities from one waste-heat stream.

Building integration: the CoolSkin façade

CoolSkin (SFB1244, Stuttgart) integrates a closed low-pressure adsorption system into the façade itself: solar heat regenerates the adsorber (reaching ~100 °C at end of regeneration), and the panel delivers 100–150 W of cooling per installed m² of adsorber façade, sustained for ~12 hours. The envelope becomes the cooling machine — the architectural endpoint of solar adsorption cooling, and the bridge to the radiative/façade and envelope threads.

Commercial reality check (note: absorption vs adsorption)

Some catalog data in the set is absorption, not adsorption — e.g. Thermax low-temperature hot-water-driven absorption chillers, 200–1640 TR (700 kW–5.76 MW), COP 0.81 from 80 °C hot water. Absorption (liquid LiBr/water) and adsorption (solid sorbent) are siblings: absorption has higher COP and power density and dominates large commercial waste-heat cooling; adsorption wins on lower drive temperature, no crystallization/corrosion, no moving liquid, and solid-state simplicity at smaller scale. For 601, the choice between them is a real down-select — absorption if the scale and a steady 80–90 °C stream justify it, adsorption for lower-grade/intermittent heat and simplicity.

What this means for 601 Delaware

• Expect COP ~0.4–0.5, SCP ~150–200 W·kg⁻¹ from a conventional silica-gel/water unit on roaster waste heat at ~80 °C; EMM-8-class aluminophosphates are the upgrade path that keeps COP high if the available heat is cooler (60–70 °C).
• Size the adsorbent mass from SCP and the roaster’s actual exhaust enthalpy/duty cycle, and derate hard for the August condenser temperature — pair with the best available rejection sink.
• The desalination cogeneration and CoolSkin façade angles are genuinely on-brand options worth a feasibility look, not just lab curiosities.

See also

• Adsorption Cooling — the conceptual overview this page quantifies
• Limitations & Mitigations — the derating curve and mitigation stack (mass/heat recovery, multi-bed, Stratisorp) behind these numbers
• Commercial Adsorption Chillers — vendors and field deployments hitting these figures
• Composite Salt Sorbents — salt-in-matrix working pairs that lift uptake 3–5× over silica gel
• Adsorbent Bed Engineering — bed geometry/coatings/TPMS that raise SCP toward these numbers
• Solar Adsorption Cooling — these figures as realized in solar-driven field systems

Full-text corroboration (2026 Zotero ingest)

Provenance for several headline figures is now backed by the full-text papers, not just registry metadata: the EMM-8 COP 0.85 @ 63 °C benchmark traces to its primary source (Liu et al. 2022, Nature Communications); Frazzica et al. (2021) screen ten working pairs (water/ammonia/ethanol/methanol refrigerants) for low-temperature waste-heat upgrading; and the coated-bed ceiling is sharpened by an automotive AQSOA-Z02/water study where a BACT coating reached ~1875 W·kg⁻¹ SCP (Jacobucci) — the high end of the SCP range and a concrete data point for Adsorbent Bed Engineering. Elsheniti (2018) corroborates the working-pairs table (zeolite 3300–4200 kJ·kg⁻¹; CPO-27(Ni) SCP 440 W·kg⁻¹; heat recovery +25 %).

• CO₂ (R744) — activated-carbon/CO₂ pairs for sub-zero/pressurized duty
• Geothermal — the heat-rejection sink the hot-day derating demands
• Heat Battery — buffering intermittent drive heat to hold COP
• 601 Delaware — Cooling Strategy — the building application
• 601 Delaware — Adsorption Chiller Sizing — these COP/SCP figures applied to size adsorbent mass against the roaster exhaust (~20–40 kg silica gel → ~1 ton)
• raw sources