District Cooling Malaysia: What You Actually Pay For

The pipes under Putrajaya

Since 1999, most government buildings in Putrajaya have had no chillers of their own. The cold arrives through a pipe.

A short distance from the ministries, a central plant makes chilled water around the clock and pumps it through insulated underground pipes to dozens of buildings. Each building draws the cooling it needs through a heat exchanger in its basement, warms the water up in the process, and returns it. That loop, one plant cooling a whole district, is district cooling, and Malaysia runs some of the largest schemes in the region.

For a building owner the deal sounds simple: stop owning chillers, start buying cold as a service, the way you already buy electricity. Whether the deal is actually good for you depends on the contract, and on a handful of numbers this article will teach you to read. We will use real Malaysian schemes, and real lessons from a district cooling plant we helped run.

Who runs district cooling in Malaysia?

The three longest-running schemes cover the administrative capital, the KLCC precinct and Cyberjaya.

Newer city districts build their own central plants too. Tun Razak Exchange (TRX) in Kuala Lumpur serves its towers and mall from a central energy plant, and large campuses such as hospitals and universities increasingly use the same model at smaller scale: one plant, many buildings.

Why does the model make sense here? Malaysia's cooling load never really goes away. There is no winter, so a central plant runs profitably all year, and buildings that share one big, well-run plant can beat the efficiency of a dozen small, half-loved chiller rooms. The catch is that "well-run" is doing a lot of work in that sentence. More on that below.

How the cold reaches your building

A district cooling scheme has three parts, and the boundary between the last two is where your money changes hands.

1. The central plant holds the chillers, pumps, cooling towers and controls. Many plants also have thermal storage: big tanks of chilled water made at night, when electricity is cheaper and the plant is quiet, then drawn down during the afternoon peak.
2. The distribution network carries supply water out to customers, typically at about 4 to 6 °C, and brings warmer return water back.
3. The energy transfer station (ETS) sits in your building. Plate heat exchangers transfer cold from the district loop to your building's own chilled-water loop. The two water circuits never mix. A billing meter at the ETS records how much cooling you took.

The heat exchanger is the practical dividing line. The operator controls everything needed to deliver cold water to that boundary. Everything after it, the secondary pumps, the air handling units (AHUs), the fan coil units (FCUs), the schedules, the tenant complaints, is still yours, and still needs a properly tuned building management system (BMS). Buildings sometimes discover this the hard way: they connect to district cooling, the plant room disappears, and the hot-office complaints stay exactly where they were, because the problem was on the air side all along. For a refresher on that side of the loop, see how a chilled-water system works.

What is on the bill: RT, RTh and capacity

District cooling bills are built from two quantities, and they behave like the two halves of a TNB maximum-demand bill.

A refrigeration tonne (RT) is a rate of cooling, like speed. One RT equals 3.517 kilowatts (kW) of cooling being delivered at this moment.

A refrigeration tonne-hour (RTh) is an amount of cooling, like distance. One RTh is 1 RT sustained for one hour.

If your building draws an average of 800 RT across a 10-hour working day, the day's consumption is:

$$800\ \text{RT} \times 10\ \text{h} = 8{,}000\ \text{RTh}$$

Over 30 such days that is 240,000 RTh for the month. A typical contract then charges you twice:

$$\text{Bill} = C_{\text{cap}} \times RT_{\text{contracted}} + C_{\text{use}} \times RTh_{\text{metered}} + \text{adjustments}$$

The capacity charge \(C_{\text{cap}}\) pays for the right to draw up to your contracted RT, whether you use it or not. It funds the operator's plant and pipes. The consumption charge \(C_{\text{use}}\) pays for the RTh you actually metered. Rates and definitions differ by scheme and contract, so never compare two proposals on one headline rate.

The capacity line deserves the most attention before you sign. Contract too much RT and you pay every month for cooling you never draw; too little and expansion gets awkward. Ask how contracted capacity is set, whether it resets from measured peaks, what the minimum-consumption clause says, and who owns and calibrates the billing meter. These questions are boring right up until the first surprising invoice.

Inside the plant: kW/RT decides the economics

Here is the part of the industry we know personally. Cobler delivered the BMS for the district cooling plant that serves EPF's headquarters, Malaysia's national retirement fund, on an MRCB-led project. The building runs every hour of the day, and the plant's job is to make cold water for it as cheaply as physics allows.

A chilled-water plant is judged on one number: kW/RT, the kilowatts of electricity it burns for every tonne of cooling it delivers. It is the plant's fuel economy. Most plants sit between 0.85 and 1.0 kW/RT. The EPF plant's brief was 0.7, and it holds it through the day, not just at the design point on paper.

The difference sounds small until you scale it. A district-sized plant delivering 5,000 RT that improves from 0.9 to 0.7 kW/RT saves:

$$(0.9 - 0.7)\ \text{kW/RT} \times 5{,}000\ \text{RT} = 1{,}000\ \text{kW}$$

That is a megawatt of electricity, every running hour, for the same cooling. It is why an efficient operator can offer a competitive tariff and an inefficient one cannot, and it is the number your own chiller room competes against if you stay independent.

Getting from 0.9 to 0.7 is not about buying one heroic chiller. Chillers have a band where they run efficiently; lightly loaded they waste energy, and flat out they waste energy. Most of the gain comes from staging: running exactly as many chillers as the moment's load needs, letting the most efficient machine carry the base, and dropping one the instant demand falls. The pumps and cooling towers then follow the real load instead of moving water for capacity that is not even switched on. The cooling demand moves all day, so holding 0.7 takes constant adjustment. That is the control system's job, and it is why the plant's brain matters as much as its hardware.

The delta-T trap

There is one way a customer can quietly damage this whole machine: low delta-T.

Delta-T is the temperature difference between the water you return and the water you were supplied. The cooling you extract from the water is:

$$Q = \dot{m} \times c_p \times \Delta T$$

where \(Q\) is cooling power, \(\dot{m}\) is water flow, \(c_p\) is water's specific heat and \(\Delta T\) is your delta-T. Take out the same heat with a smaller delta-T and the flow must rise to compensate. A building designed for a 6 °C delta-T that operates at 4 °C needs:

$$\frac{6}{4} = 1.5 \times \text{the design water flow}$$

for the same cooling. Half again as much water, pumped through the same pipes, using network capacity the operator planned to sell to someone else. Low delta-T usually comes from mundane faults on the customer side: control valves stuck open or leaking, oversized bypasses, fouled coils, a sensor reading wrong, or AHUs pushed past their coils' ability to absorb heat. The water comes back cooler than intended, which sounds harmless and is not.

Operators know this, which is why many contracts carry a low delta-T clause: a penalty rate, a flow limit, or a minimum return temperature. If you connect to district cooling, your delta-T becomes a bill item, so it pays to know what causes low delta-T syndrome and how to fix it.

Should you connect, or keep your own chillers?

There is no universal answer; there is a trade.

Compare on whole-life cost, not on year one: include the ETS, internal distribution, chiller replacement cycles you avoid, the rentable value of a freed-up plant room, and the contract's escalation and minimum-take terms across its full tenure. Then stress-test more than one load profile. A tower that contracted generous capacity and then lost an anchor tenant is paying capacity charges on empty floors.

Watch the boundary meter

Whichever side of the trade you are on, the ETS meter is where the money flows, and it deserves independent eyes.

Record supply and return temperature, flow, instantaneous RT and accumulated RTh at the boundary, separately from the operator's billing system where practical. Do not expect two meters to agree at every five-minute interval; align timestamps, integration periods and meter constants first, then compare daily and monthly totals against the contract's meter tolerance. Trend your delta-T against load, because a drifting delta-T is both an early fault alarm and a future bill item. And when a bill jumps, data at the boundary tells you whether cooling use genuinely rose or a meter needs a conversation.

This is the layer CobiNeural covers: metering at the ETS, chilled-water and energy trends, delta-T and kW/RT tracking, alerts, and bill verification, running standalone or on top of the BMS, PLC or SCADA already in the building. We watch the same numbers on the plant side at EPF HQ, so we know what healthy looks like from both ends of the pipe.

If you are weighing a district cooling contract, or trying to make sense of one you already have, talk to us. Bring last month's bill; it is usually an interesting conversation.