Joule Heist search

topics · updated 2026-09-25

Solar Adsorption Cooling

confidence: high volatility: warm verified: 2026-09-25fresh

Solar-driven adsorption cooling — the load-coincident application where sunlight supplies the desorption heat. Spans intermittent single-bed solar ice makers (the collector IS the adsorber; carbon/methanol classically, carbon/ethanol now the pair to beat) to two-bed continuous silica-gel/water chillers with storage. Field and simulation results cluster at cyclic COP ~0.35–0.5 and solar COP ~0.1–0.3; silica-gel/water regenerates at 60–95 °C, needs ~3.5 m² of collector per kW, and collector area and cost are the dominant economic levers. Mass/heat recovery, multi-bed operation, buffer storage, and cold heat-rejection sinks are the recurring performance boosts; maintenance and sorbent durability decide long-term output.

The most on-brand application of Adsorption Cooling: use sunlight as the desorption heat, so peak supply (midday sun) roughly coincides with peak cooling demand. This article maps the design space — from intermittent single-bed solar ice makers (the collector is the adsorber) to two-bed continuous silica-gel/water chillers with thermal storage — and the measured/simulated performance across climates. The recurring lessons: cyclic COP ~0.35–0.5, solar COP ~0.1–0.3, and collector area + cost dominate the economics. See Solar Thermal for the collectors themselves.

The cycle in one paragraph

Every solar adsorption machine runs the same four-phase closed cycle (Al-Yasiri; Hlail): heating/pressurization → heating with desorption (vapor condenses) → cooling/depressurization → cooling with adsorption (refrigerant boils in the evaporator, producing the cold). The bed is the “thermal compressor”; there is no compressor and, in the simplest builds, no valves. One bed produces cold only in the last phase, so continuous cooling needs two or more beds out of phase; a single bed gives one cold pulse per day. Adsorbents: activated carbon, silica gel, zeolite; refrigerants: water, methanol/ethanol, ammonia (pair selection in Adsorption Cooling).

The two architectures

Intermittent (single-bed) solar ice makers. The simplest form: a flat-plate collector whose absorber plate is the adsorbent bed. By day the sun desorbs the refrigerant (condensed and stored); by night the bed cools and re-adsorbs, producing the cooling effect — one cycle per day. Working pairs are usually activated carbon/methanol or activated carbon/ammonia, and since 2026 activated carbon/ethanol (below). Documented examples:

  • A coconut-shell activated-carbon/methanol valveless, gravity-fed ice maker: 1 m² stainless-steel bed (19 kg carbon, internal fins, a “false bottom” for vapor distribution) solar-heated to 90–100 °C under a selectively coated cover, with condenser and evaporator below it (Dhokane). Stainless was chosen over aluminium because aluminium catalyzes methanol decomposition — a detail every carbon/methanol builder should know.
  • A BSc-scale AC/methanol ice maker at Palestine Polytechnic aimed at off-grid food and vaccine storage, whose literature review notes that early zeolite–water solar work (Tchernev; Pons & Guilleminot) reached COP of only ~0.1 — the motivation for choosing carbon/methanol (Hlail).
  • A hybrid charcoal/methanol + charcoal/R-11 prototype (IUT Bangladesh) that recovers hot water during the adsorption half-cycle as a co-product; the full system leaked, and a modified bench prototype (500 g charcoal, electric heater as stand-in sun) reached evaporator minima of 17 °C (methanol, ~119–300 mmHg) and 16 °C (R-11, 1.7–4.2 bar) (Rahman). Cooling, not ice — and a reminder that leak-tightness kills first prototypes.

Cheap, unpowered, silent — but low COP and once-daily.

Continuous two-bed chillers. Two sorbent beds alternate adsorption/desorption so cooling is continuous, driven by hot water from collectors (usually silica-gel/water, chosen because silica gel regenerates at a lower temperature than zeolite or carbon and water has a high latent heat). Bhargav’s review gives the working envelope: regeneration ~70–80 °C, chilled water ~10–20 °C, system pressure 1–12 kPa, chiller COP 0.3–0.5. These are the building-scale machines — the same hardware sold commercially (see Commercial Adsorption Chillers), just solar-fed. Concentrating variants (parabolic trough or CPC onto a tubular bed) reach the higher temperatures that zeolites want (Yuan).

Measured & simulated performance

StudySystemClimateKey numbers
Alahmer (Perth)Two-bed silica-gel/water + CPC collectors, dynamic modelTemperate, high-DNI; summer 30–40 °C~11 kW at 13:00; cyclic COP ~0.5, solar COP ~0.3; optimum ~38 m² CPC, ~11 yr payback
Bawazir (Riyadh)TRNSYS, 80-villa silica-gel/water, ETC + storage, 60–95 °C driveHot desert, 6.0 kWh/m²/day96 % solar fraction at 5500 m² ETC + 350–400 m³ storage; 74 % energy-cost cut, ~75 % CO₂ cut vs vapor compression
Najeh (Nancy)Two-bed silica-gel/water (ENERBAT), 5.2 m² flat plateContinentaldaily COP 0.35–0.37; machine starts at 74 °C tank temperature; collector output up to 94 °C; geothermal-cooled adsorber roughly halves cycle time vs ambient-air cooling
Bhargav review (El-Sharkawy)CPC two-bed silica-gel/water, 900 s cycleCairo/Jeddah/Aswancycle/solar COP 0.50/0.30; 14.8–15.8 kW; hot-water buffer adds up to +12 % daily capacity, +9 % COP
Bhargav review (Chang, Taiwan)Solar compound system, plate-fin bedsSubtropicalCOP 0.37–0.40, 9 kW; solar-heating η 28 %, cooling 45 %, solar cooling η 12.8 %
Yuan (Beijing)SAPO-34 vs ZSM-5 zeolite/water, tracked parabolic trough (2.4 m²) on a black-chrome tubular bed—SAPO-34 wins: COP 0.169 (1800 s adsorption, SCP 79 W·kg⁻¹) or SCP 169.7 W·kg⁻¹ (600 s, COP 0.122) — COP and SCP trade off with adsorption time
Dias (Aveiro)Silica-gel RD2060/water, 50-tube cylindrical bed, 2-D modelWarm summer dayCOP 0.5, SCP 44 W·kg⁻¹ at 70/30/15 °C; hot day (42 °C condenser) → COP 0.31, SCP 26; refrigeration (2 °C evaporator) → COP 0.25, SCP 18

The pattern: two-bed silica-gel/water lands around cyclic COP 0.35–0.5; solar COP (which folds in collector efficiency) is lower, ~0.1–0.3 — Bhargav quotes 0.3–0.5 for the chiller and 0.1–0.3 as the realistic optimum for solar-driven systems, which is the same distinction, not a disagreement. Concentrating/intermittent variants trade COP for simplicity or higher drive temperature. Al-Yasiri’s compilation of best cases runs higher (silica-gel/water COP 0.63 at 85 °C; SAPO-34/water 0.575 at 75 °C; carbon/ethanol 0.68 under a parabolic trough), but these are peak laboratory points, not daily field averages.

What moves the numbers (Bhargav; Al-Yasiri; Dias):

  • Colder heat rejection. Lower cooling-water temperature raises both COP and capacity; the hot-day condenser derating from Limitations & Mitigations applies directly (Dias’ 42 °C case drops COP from 0.5 to 0.31). Najeh’s geothermal-cooled adsorber is one mitigation, halving cycle time.
  • Warmer chilled water. Air-conditioning duty (12–18 °C chilled water) performs far better than refrigeration duty (Dias: COP halves at a 2 °C evaporator).
  • Mass and heat recovery. A mass-recovery step (~60–180 s) plus heat recovery between beds raises COP ~25 % (Wang et al. in Al-Yasiri), and up to +100–125 % with pre-heating/pre-cooling on a zeolite-composite cycle (Tso et al.); a GA-optimized recovery schedule gained 8.3 % capacity on a 450 kW unit (Rezk & Al-Dadah in Bhargav).
  • Multi-bed operation. Four beds vs two: ~12 % higher SCP at the same heat-exchanger inventory and ~25 % higher COP (Wang et al. in Bhargav).
  • Regeneration temperature. Raising it lifts SCP much more than COP (Dias; Rezk: 40→65 K lift nearly doubles capacity at a COP cost of only 0.63→0.60). Caps differ by source: Yuan advises keeping silica-gel desorption below ~95 °C to avoid dehydration damage, Motamedi caps it at 120 °C with ~90 °C nominal, Bawazir’s design runs 60–95 °C — treat ~90 °C as the safe working ceiling for silica gel.
  • Cycle time. SCP and COP move oppositely past an optimum; each system needs its own cycle time (Dias: reference 22 min; Yuan: two-thirds of cooling delivered in the first 600 s of adsorption).
  • Buffer storage smooths solar intermittency and raised daily capacity up to 12 % in El-Sharkawy’s CPC system (Bhargav); at neighborhood scale (Bawazir) 350–400 m³ of storage is what makes a 96 % solar fraction possible.

Economics: collector area is the lever

Across the reviews and techno-economic studies the same conclusion recurs: the solar collector field dominates both cost and footprint.

  • Collector area per kW. Adsorption needs the largest collector field of the closed sorption cycles — ~3.5 m²/kW versus ~2.8 m²/kW for absorption and 1.3–1.7 m²/kW for desiccant systems (Henning, in Al-Yasiri). Alahmer’s Perth optimum of ~38 m² for ~11 kW is consistent.
  • Small scale is the cost-efficient scale. A particle-swarm-optimized single-family design (silica-gel/water, flat-plate collectors at 221 US$/m², adsorption unit at 700 US$/kW) found the optimum ~13 % cheaper than the baseline — mainly by halving collector area (20→10 m²) and tank volume while raising hot-water flow — with the minimum cost per kW in the 5–10 kW band and acceptable up to ~50 kW; the collector’s initial cost was the dominant bottleneck, and with very cheap electricity the optimizer preferred auxiliary electric heat over more solar. A “green” 50 m²/10 kW design cost 21 % more per year but avoided ~15.7 t CO₂ (Motamedi).
  • Large scale pays where cooling dominates the grid. In the Gulf, cooling can exceed 70 % of electricity use; Bawazir’s 80-villa Riyadh system (up to 2 MWth) cut annual energy cost 74 % and CO₂ ~75 % against vapor compression, assuming a 25-year chiller life vs 15 for the VCC.
  • Maintenance is the quiet advantage. Jordan’s economic study contrasts an adsorption chiller (municipal water + silica gel, up to 8000 h/yr, full capacity within 8 min of a restart, ~$5,000/yr maintenance) with a LiBr absorption chiller (daily dilution, monitoring, exchanger replacement, ~$20–30k/yr); its RETSCREEN scenarios for Aqaba pair a two-stage adsorption chiller with a CPC evacuated-tube field or waste heat, a dry cooler or heat-recovery sink, and a downsized VCC for peak load (Saidi).
  • Market position. Absorption still dominates installed solar cooling (~59–60 % of systems; adsorption ~11 %), reflecting its higher COP (single-effect ~0.7 at 85 °C, double-effect 1.2 at 130 °C) and smaller collector field. Adsorption’s case rests on lower regeneration temperature, no crystallization or corrosion, no moving parts, and low maintenance (Al-Yasiri; Saidi). Solid-desiccant systems are a third option for humid climates (Al-Yasiri).

Favorable cases concentrate in high-DNI, high-cooling-load climates — Riyadh, Aqaba, Perth, the Gulf and the San Antonio profile.

2025–2026: carbon–ethanol displaces carbon–methanol; the field takes stock

Three additions from the August 2026 research round move this article’s guidance:

  • Maxsorb III/ethanol doubled the classic ice maker. A 2026 field study in Mosul, Iraq (Soheel et al., 85 field experiments over a semi-arid summer — rare replication volume for this literature) ran a 1 m² intermittent single-bed ice maker on Maxsorb III superactivated carbon (BET 3045 m²/g, pore volume 1.70 cm³/g) + ethanol. Best case, at a low charge (0.8 kg ethanol per kg carbon) and 3 kPa evaporator pressure: 0.71 kg ethanol cycled, 2.1 kg ice/day, evaporator sub-zero for 5.8 h, maximum COP 0.583 — +110 % ice yield and +35.6 % COP versus a NORIT PK1 carbon–methanol benchmark of the Pons/Guilleminot lineage (see the DIY reference for that baseline). Caveat: the 0.583 appears to be a cycle/thermal COP, not a solar COP; the classic ~0.10–0.16 solar-COP band still applies for cross-comparison until the full text is checked. Ethanol also sidesteps methanol’s toxicity and its >120 °C decomposition failure mode. The note records that a University of Dayton review (Choi et al., 2025) independently names carbon–ethanol the emerging pair of choice. For a new intermittent build, carbon–ethanol is now the default to beat.
  • The Dec 2025 state-of-the-field review (Patel et al., open access) synthesizes ~30 years of solar adsorption work: cycle COP band 0.2–0.6 (matching the table above), system cost and zeolite cycling degradation as the adoption barriers, ≥50 % electricity reduction in arid-zone case studies, and — for long-life field systems — ranks zeolite 13X–water the most practical pair overall on heat tolerance and lifespan, which supports the safe zeolite–water demonstrator path in the DIY reference. It also elevates maintenance (collector cleaning, adsorbent monitoring and replacement) to a first-order determinant of long-term output, and points future work at MOF/composite sorbents, hybrid cycles, and techno-economic/policy analysis.
  • The honest headwind, quantified. A 2026 Scientific Reports head-to-head (Selvaraj et al.) drove the same 60 W sorption refrigerator from a flat-plate collector and from a 100 W PV panel + battery under Indian tropical conditions: the thermal route gave COP 0.14 at 35.3 % collector-to-generator efficiency (generator peak 88 °C — squarely the flat-plate band assumed above) with stable output; the PV route gave COP 0.12 at ~9 % end-to-end efficiency, a colder 9 °C evaporator, and cloud-sensitive swings. The machine was an ammonia–water absorption unit, so transfer qualitatively. Routing sunlight through heat into a sorption machine is ~4× more efficient end-to-end than routing it through electrons, but PV+vapor-compression’s system COP still wins the market — see PV vs Thermal for the 2025–26 sharpening of that trade.

See Also

Sources