Inside a 23,300 kW Propylene Brine Chiller Package for an Acrylonitrile Plant

Chinese acrylonitrile capacity crossed 5.7 million tonnes per year at the end of 2025, after more than 1.3 million tonnes of new capacity came online in a single year. Downstream acrylonitrile-butadiene-styrene resin and acrylic-fibre operating rates have been stuck in the 60–70% band for several quarters. Commodity analysts at Chinese market intelligence house SunSirs summed up their January 2026 annual review with the phrase “a continuation of the darkest hours, and a shakeout.”

For EPC contractors and process licensors working on acrylonitrile projects, that market picture has a very practical consequence: capital and operating cost scrutiny has moved from the reactor and the recovery train outward into the utility blocks. Refrigeration — historically treated as a package to be dropped into the utility scope late in the design — is now being interrogated at basic-engineering stage.

This article works through one reference point from that world: a pair of 23,300 kW propylene-refrigerant brine chiller packages that Suzhou Lmart Energy Equipment supplied in 2022 for an acrylonitrile plant operated by Sinopec Zhenhai Refining & Chemical. It remains the largest single refrigeration duty on our project list, and it is a useful anchor for anyone sizing an ultra-high-capacity chilled-brine system.

Why acrylonitrile recovery depends on absorber chilling

Commercial acrylonitrile is made by the ammoxidation of propylene: propylene, ammonia and air react over a fluid-bed catalyst, and the hot reactor effluent is quenched before entering an absorber, where chilled water strips acrylonitrile out of the gas phase for downstream purification.

The physics is unforgiving in a helpful way. The colder the absorber water, the more completely acrylonitrile is captured, and the less product leaves with the vent gas. Because the absorber operates near ambient conditions on the gas side, the whole recovery efficiency of the front end is effectively set by how cold, and how reliably cold, the circulating water can be held.

Almost every modern design therefore places a brine refrigeration system on the absorption section: a chiller package cools an ethylene-glycol/water solution to the design supply temperature, and that secondary coolant carries the duty to the absorber and to any other cold users in the block. The secondary-loop arrangement is chosen deliberately — it keeps the flammable primary refrigerant inventory confined to the machinery area, tolerates dispersed cold consumers, and gives freeze protection margin.

What has changed is the commercial weight of that duty. In a strong market, a tenth of a percent of recovery is a rounding error. At current operating rates, with much of the industry at thin margins, it is cash flow — and the guaranteed capacity and part-load efficiency of the refrigeration package end up written into the plant's cost model.

Single-train acrylonitrile capacity has also grown steadily, with 130,000 and 260,000 t/a lines now routine. Scale up the plant and you scale up the cold duty, which is how refrigeration packages in this service moved from the low thousands of kilowatts into the tens of megawatts.

Project overview: 23,300 kW brine chiller package

The headline parameters of the reference project are set out below.

End user	Sinopec Zhenhai Refining & Chemical
Process unit	Acrylonitrile plant, absorption section
Package type	Brine chiller (secondary coolant loop)
Refrigeration duty	23,300 kW (23.3 MW)
Refrigerant	Propylene (R1270)
Secondary coolant	Ethylene glycol / water solution
Quantity	2 units
Delivery	2022

Lmart's role on this project was equipment supply: we designed, manufactured and package-integrated the chiller units and supplied them to the plant. Process design, plant integration and commissioning of the acrylonitrile unit itself sat with the owner and its engineering contractor.

Three engineering decisions behind the package

1. Propylene as the refrigerant

Selecting propylene (R1270) rather than a synthetic refrigerant is the standard call in this service, and the reasoning is more operational than thermodynamic. The plant produces propylene as its own feedstock, so make-up refrigerant comes from an existing on-site stream rather than a procurement channel. The owner's operators and the design institute already handle the fluid daily and know its property data. On top of that, propylene has a high volumetric refrigerating capacity and good low-temperature behaviour, which keeps swept volume — and therefore machine size — within sensible bounds at this duty.

The cost is that the package carries a flammable inventory. Hazardous-area classification, gas detection, purge and pressure-relief arrangements, and the electrical specification of every motor and instrument on the skid all follow petrochemical practice rather than commercial HVAC practice. In practical procurement terms this is a filter: it separates suppliers with genuine hydrocarbon refrigeration experience from those extrapolating from lighter duties.

2. What 23,300 kW actually means for the machinery

A useful scale check: a domestic split air conditioner is around 3.5 kW of cooling. This package is the equivalent of roughly 6,600 of them — serving the absorption section of a single process unit.

At that duty, the design cannot be a scaled-up small machine. Compression is arranged as multiple oil-injected screw compressor packages rather than a single machine, so that the plant retains capacity during maintenance and holds efficiency across the load curve. Evaporators and condensers become large pressure vessels in their own right, frequently split into parallel shells; any maldistribution between those shells shows up directly as lost design margin. Oil separation and oil cooling scale non-linearly with compressor size and become a genuine design task rather than a selection. And the control philosophy — capacity staging, sequencing, anti-surge of the coolant loop, safe shutdown of a flammable-refrigerant system — is written for the project rather than pulled from a standard library.

3. Matching the secondary coolant loop

The most common cause of a chiller package under-delivering at the plant boundary is not the machine. It is the glycol loop: concentration, flow rate and supply/return temperature difference that do not match the absorber's heat transfer duty. A package can meet its test-bench rating perfectly and still leave the process short of cold.

Our practice is to engineer the evaporator together with the package rather than buy it as a component. Lmart manufactures its own shell-and-tube heat exchangers to TEMA practice and to ASME BPVC Section VIII requirements, which means the surface margin, tube-side velocity and shell-side flow arrangement are visible and negotiable during design review rather than buried in a vendor's sub-order.

Track record at the 10 MW-plus level

Ultra-high-capacity refrigeration is an experience-dense product, so the more useful question for a buyer is not “can you build it” but “what have you built at this duty, with this refrigerant.” The comparable deliveries on our list are set out below.

Year	End user	Application	Duty	Refrigerant
2022	Sinopec Zhenhai Refining & Chemical	Acrylonitrile	23,300 kW	Propylene
2019	Zhejiang Petroleum & Chemical	Acrylonitrile	14,400 kW (3,500 kW motor)	R1270
2022	Shandong Yulong Petrochemical	Acrylonitrile	11,849 kW	R1270
2022	Sinopec Quanzhou	Petrochemical process cooling	14,000 kW	Propylene
2022	Sinopec Zhenhai Refining & Chemical	Propane dehydrogenation (PDH)	8,330 kW at −32°C	Propylene

Three acrylonitrile brine packages, all on propylene-family refrigerant, all with glycol/water as the secondary coolant. That is not coincidence — it is the combination that has been repeatedly validated for this process, and it is a reasonable starting assumption for a new project unless something specific rules it out.

Six selection criteria for ultra-high-capacity chiller packages

Drawn from the projects above, here is what we would put on a technical bid tabulation for a package in this class.

1. Let the plant choose the refrigerant. In a petrochemical complex, a hydrocarbon refrigerant already produced on site (propylene, propane, ethylene) usually wins on make-up logistics and operator familiarity. Reserve synthetic refrigerants for sites with a hard constraint on flammable inventory, and keep an eye on the direction of GWP regulation over the asset's life.
2. Compare multi-machine and single-machine arrangements over the whole life, not at bid price. A parallel arrangement usually costs more up front. It repays through better part-load efficiency and the ability to maintain one machine without shutting the process down — value that typically appears from year three onward on a continuously operating plant.
3. Treat the heat exchangers as primary scope. At this duty the evaporator and condenser are large pressure vessels with real tube-vibration, flow-distribution and fouling-margin questions. Whether the package supplier manufactures them determines how short the responsibility chain is when performance is queried.
4. Specify the degree of skid integration explicitly. Machinery, oil system, instrumentation and piping assembled and tested in the works turns site work into lifting and tie-ins. On a schedule-driven project the site hours saved routinely exceed the price difference between a loose-supply and a skid-integrated offer.
5. Demand like-for-like references. Ask for a project list filtered by duty band, refrigerant and secondary coolant, with end user and year, and check it line by line. Experience at 1 MW does not transfer to 23 MW.
6. Fix the certification basis early. Pressure parts built to ASME BPVC Section VIII versus PED 2014/68/EU versus a national code carry different design, inspection and documentation routes. Deciding at technical-agreement stage rather than after award protects the manufacturing schedule and the third-party inspection window.

What Lmart manufactures, and what it integrates

Being explicit about scope avoids a common misunderstanding in this equipment class.

Manufactured in-house: pressure vessels, shell-and-tube heat exchangers (evaporators, condensers, oil coolers, economisers), and the mechanical package and skid integration itself — structural skid, piping, oil system, instrumentation, control panel, assembly and works testing.

Bought in and integrated: the compressor. Lmart does not manufacture compressors. We build our refrigeration packages around bought-in oil-injected screw compressors, and where a project calls for a centrifugal or other machine type we buy that in and integrate it the same way. What we are responsible for is the unit — selection, thermal and mechanical design around the machine, the heat transfer equipment, the skid, the controls and the tested package.

Suzhou Lmart Energy Equipment Co., Ltd. is a member of the Huachang Group. Since 2016 we have delivered more than 70 industrial refrigeration and gas compression packages into the petrochemical sector — acrylonitrile, PDH, polyolefins and ethylene tank farms among them — with duties from 45 kW to 23,300 kW, alongside a pressure vessel and heat exchanger business exporting to over 30 countries.

If you are sizing a high-capacity chilled-brine system, or benchmarking suppliers for one, our engineering team can return a technical proposal with a general arrangement drawing and budget pricing within 48 hours. Get in touch with the duty, refrigerant, coolant temperatures and site conditions and we will work from there.