The cold-storage industry has more or less finished its move off HFCs. The question of what to replace them with has split into two camps. Ammonia (R-717) and CO2 (R-744), sometimes in cascade with ammonia, sometimes in transcritical configurations. Both are natural refrigerants with zero global warming potential, both have decades of operating data, and both have distinct trade-offs that matter on the operating floor.
At Colden we run ammonia at our Mehsana facility, with a small CO2 cascade scoped for the blast freezer in a 2027 expansion phase. We have run ammonia plants long enough across the team's prior careers to have opinions. Here is the comparison as of 2026.
Ammonia, the long-standing standard
Ammonia is the refrigerant cold-storage has run on since the 1880s. It has the best thermodynamic efficiency of any refrigerant in commercial use. Operating cost per pallet held is the lowest, typically 6 to 12 percent lower than CO2 systems running the same duty.
The downsides are safety and regulation. Ammonia is toxic at low concentration. Ammonia plants in India come under PESO and MAH classification when the charge exceeds threshold quantities, which adds documentation, training, and inspection burden. Leaks must be reported under state pollution control board rules, and personnel must be trained on response.
In practice these downsides are manageable. The industry has 140 years of safety procedure. Modern plant rooms are gas-tight, sensor-equipped, and engineered to limit leak migration. The regulatory burden is real but not onerous for a properly run operation.
We run ammonia at our Mehsana facility. The decision was straightforward: lowest operating cost, well-understood safety profile, deep field-service network in Gujarat.
CO2, the new contender
CO2 systems come in two main flavours. Transcritical CO2 runs the high side above the critical point, which works well in cold climates and has efficiency penalties in hot climates. Cascade systems use CO2 on the low-temperature side and ammonia on the high-temperature side, capturing efficiency benefits of both.
CO2 has three real advantages over ammonia. Safety profile is friendlier (CO2 is heavier than air and disperses without toxicity). Regulatory burden is lower. And the high operating pressures allow smaller pipe sizing, which reduces material cost on greenfield builds.
The downsides are operating cost and field-service depth. Transcritical CO2 in hot climates like Mehsana can run 15 to 25 percent less efficient than ammonia in the same duty. The field-service network for CO2 in India is thinner than for ammonia, especially outside major industrial estates. And the higher operating pressures require more careful component selection and maintenance.
We are scoping a small CO2/ammonia cascade for the second-phase blast freezer at Mehsana, because cascade at the blast end captures useful efficiency on rapid-cycle duty. The main room stays on ammonia.
The 2026 comparison
Here is how the two refrigerants compare on the metrics that matter for operators making a decision today.
Operating cost per pallet per month. Ammonia wins by 6 to 12 percent in most Indian climates. CO2 closes the gap in cold climates and loses it in hot ones.
Capex per cu ft. Ammonia wins on greenfield builds where the plant room can be designed cleanly. CO2 closes the gap in tight retrofits where pipe sizing matters.
Safety and regulatory burden. CO2 wins. Ammonia is manageable but not free.
Field-service network. Ammonia wins outside major metros. CO2 is closing the gap in major metros but not in tertiary markets.
Carbon footprint of the refrigerant itself. Tie. Both are essentially zero GWP.
Carbon footprint of the operating electricity. Ammonia wins where it operates more efficiently. CO2 wins in cold climates where the climate favours it.
Future-proof against regulation. Tie, at least for the next decade. Both are off the HFC phase-down list.
A practical decision framework
For most Indian cold-storage operators making a greenfield decision in 2026, the right answer is ammonia, unless one of three conditions applies.
Condition one: the local regulatory environment is hostile to ammonia. In a few jurisdictions, ammonia permitting has become slow and uncertain. Where that is the case, CO2 cascade is a reasonable alternative.
Condition two: the climate is cold enough that CO2 efficiency matches or beats ammonia. The Indian climate almost never favours CO2 on this metric, although the Himalayan belt comes closest.
Condition three: the operating philosophy values lower regulatory burden over lower operating cost. Some operators choose CO2 specifically to avoid the PESO and MAH overhead. This is a legitimate trade-off, particularly for smaller operators without a dedicated safety team.
For retrofit decisions, the calculation is different. Retrofitting an existing HFC plant to ammonia is usually cheaper than retrofitting to CO2, because the plant room layout and piping geometry are closer to ammonia conventions. Retrofitting to CO2 typically means a more extensive plant room rebuild.
Where we expect the industry to land
By 2030 we expect the Indian cold-storage industry to be roughly 70 percent ammonia, 20 percent CO2 cascade, and 10 percent transcritical CO2 in cold climates, with a long tail of legacy HFC plants still running. By 2035 the HFC tail will be largely gone, with the ammonia and CO2 shares stabilising in the same ratio.
The point is that the conversation is no longer "natural vs HFC." That is settled. The conversation is "which natural refrigerant fits this site, this climate, this regulatory environment, and this operating philosophy." That is a more interesting conversation, and the answer is genuinely different per site.