iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Malaysia’s data-centre boom is no longer a niche commercial-development story. It is becoming a national electricity-planning issue and, especially in Johor, a water-allocation issue. New facilities can add very large, steady electrical loads and substantial cooling-water demand in locations where reserve margins and treatment capacity are limited. The consequences for households depend on how quickly grid and water infrastructure is expanded, how efficiently projects operate, and whether developers pay the costs created by their connections.
The effect is not automatically a nationwide shortage or an immediate increase in household bills. It is a conditional risk: pressure is greatest where projects cluster, where peak demand coincides with other users, and where facilities rely on treated potable water rather than recycled, rain or seawater sources.
How much electricity will Malaysian data centres use?
The clearest national estimate comes from the Energy Transition and Water Transformation Ministry (PETRA), using electricity-planning figures for 2026–2035. The numbers describe a rapidly expanding pipeline rather than a single, fully operating fleet, so planned capacity, maximum demand and actual consumption must not be treated as interchangeable.
Free tools Windows power users keep installed
One-click scans. No signup required.
| Measure | Reported figure | What it means |
|---|---|---|
| Operating centres supplied by Tenaga Nasional Berhad (TNB) | 36 centres; about 4.5 GW of planned supply (Q1 2026) | Connection commitments associated with operating sites and their planned expansion. |
| Projects under construction | 23 projects; 3.8 GW maximum demand (Q1 2026) | Potential peak load once those projects are energised; it is not the same as annual energy consumed. |
| Malaysia-wide peak demand | 21.3 GW in 2026, projected to 33.5 GW in 2035 | The total system peak that generation, transmission and distribution must cover. |
| Data-centre share of electricity demand | 7% in 2026; up to 73,274 GWh, or 31%, in 2035 | PETRA’s projection of annual energy demand attributable to data centres under the stated planning scenario. |
| Separate 2025 count | 18 centres with at least 800 MW of demand in February 2025; 81 centres projected by 2035 | A different counting threshold and date, so it should not be added to the TNB totals. |
Two features matter more than any single headline percentage. First, data centres run continuously, giving the grid a high load factor but leaving less spare capacity at all hours. Second, clusters can create local transmission and substation bottlenecks even when national generation appears adequate. A project’s location and connection schedule therefore matter as much as its nameplate capacity.
Why a large load changes grid planning
Utilities must build generation, substations, transformers and high-voltage lines before a facility reaches full occupancy. They also have to manage the timing of commissioning: several campuses starting at once can create a step change in demand rather than the gradual growth associated with ordinary commercial buildings. PETRA’s figures show why multi-year forecasts now include data centres as a distinct planning category.
TNB’s response combines network investment, ultra-high-voltage tariff structures, renewable-energy procurement and efficiency conditions such as power-usage effectiveness (PUE). PUE compares a facility’s total electricity use with the electricity delivered to its computing equipment; a lower value indicates less overhead for cooling and other infrastructure. The government approved RM43 billion of TNB capital expenditure under the fourth rolling plan for 2025–2027, but that investment is for the wider power system, not a guarantee that every proposed campus can connect immediately.
Where is water pressure most acute?
Cooling design determines whether a data centre is a modest municipal-water customer or a major industrial one. Reported figures span a wide range because they refer to different facility sizes, technologies and accounting methods.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstall| Reference | Water figure | Qualification |
|---|---|---|
| Google data-centre average, 2021 | About 450,000 US gallons per day (1.7 million litres per day, or 1.7 MLD) | A company-reported average reproduced by the Malaysian Investment Development Authority and The Edge; it is not a national average. |
| Typical 100 MW data centre | About 1.1 million gallons per day (4.2 MLD) | A 2024 reported typical value; actual use varies with cooling efficiency and climate. |
| Johor data-centre requests | Average 26 MLD each year through 2036 | SPAN chairman Charles Santiago said this was additional demand above normal growth. |
| Ranhill SAJ (RSAJ) applications, 2024–2035 | 439.67 MLD requested; 316.38 MLD supported; 30.51 MLD unsupported in the cited reporting | Status categories reported by RSAJ; the figures represent applications, not water already consumed. |
| Reported reserve margins | Johor 11.5% (2022); Selangor 15.34% | Margins cited in 2024 reporting and subject to subsequent changes in supply and demand. |
The 4.2 MLD figure should be read as an order-of-magnitude indicator for a 100 MW facility, not a design standard. Air-cooled or hybrid systems can use much less potable water than evaporative systems, while hot weather, high computing utilisation and water-treatment requirements can increase withdrawals. “Withdrawal” also differs from “consumption”: water discharged after treatment may return to the environment, whereas evaporation and process losses are consumed.
Why Johor receives most of the attention
Johor has attracted a dense pipeline of large facilities near Singapore-linked fibre routes and industrial estates. Its normal annual water-demand growth was reported at about 50 MLD. The additional data-centre requests averaging 26 MLD a year therefore represent a material increment, described by Santiago as “an additional 1.3% per year above the normal organic growth.”
Applications are not the same as approved allocations or operational demand. Treatment plants, raw-water sources, pipelines and connection works may be delivered in stages, and some projects can be delayed, redesigned or cancelled. Nevertheless, the reserve margins reported for Johor and Selangor show why a cluster can become a local reliability concern before it becomes a national shortage.
Johor Bahru City Council mayor Datuk Mohd Noorazam Osman summarised the local priority this way: “People are too hyped up about data centres nowadays, but the issue in Johor is water and power.” The practical question is whether new supply and conservation projects are commissioned ahead of each campus, not whether every announced project ultimately proceeds.
Recommended Free Tools
Will data centres cause water shortages?
They can intensify scarcity risk in a constrained basin, but the available figures do not support a blanket claim that Malaysia will run out of water because of data centres. Risk is highest when four conditions coincide:
Rank #3
- Several high-water-use facilities draw from the same treatment zone.
- Reserve margins are already narrow during droughts or peak seasonal demand.
- Projects depend on treated potable water instead of non-potable or recycled supplies.
- New treatment and pipeline capacity lags the commissioning schedule.
Authorities can reduce that risk through staged approvals, binding water-efficiency requirements, alternative sources and leakage reduction. The SPAN Water Supply Guidelines for Data Centre Development took effect on 1 September 2025 and cover planning through operation, giving regulators a framework to assess a project before it receives a full supply commitment.
Will households pay higher electricity or water bills?
Government policy says data-centre developers should pay the additional cost of strengthening the electricity and water networks needed for their connections. Deputy Minister Akmal Nasrullah Mohd Nasir described this as a “user pays” approach under the new tariff structure: operators bear additional grid-strengthening costs rather than shifting them to ordinary customers. PETRA has also said domestic water tariffs will not be raised solely to fund data-centre demand.
That protection does not make impacts on household bills impossible. Electricity tariffs can change for broader fuel, generation or policy reasons, and a network investment eventually becomes part of the regulated asset base. Water utilities still face costs for dams, treatment works, pumps and leakage control. The specific safeguard is that costs attributable solely to a data-centre connection should be recovered from that user, not automatically spread across domestic accounts.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Residents may also experience indirect effects: construction can compete for land and water, and a local utility may impose connection conditions or conservation measures during tight periods. Those are service-planning issues, not evidence of an automatic surcharge on every household bill.
Is Malaysia’s grid ready for the boom?
Malaysia is preparing, but “ready” depends on the project, location and commissioning date. National forecasts include substantial headroom-building, and TNB is planning multi-gigawatt connections. At the same time, a pipeline measured in gigawatts can outpace a substation, transmission corridor or generation project in a particular state.
Signals of preparation
- RM43 billion in approved TNB capital expenditure for 2025–2027 supports generation and network reinforcement.
- Ultra-high-voltage tariffs and connection studies are being used to price large industrial loads more directly.
- Renewable procurement and efficiency requirements are intended to limit emissions and reduce avoidable electrical overhead.
- PETRA’s demand outlook explicitly models data centres through 2035 rather than treating them as incidental commercial demand.
Remaining constraints
- Peak demand can rise faster than annual energy averages suggest.
- Transmission and substation construction requires land, approvals and long lead times.
- Renewable generation needs firming, storage or flexible generation when data-centre loads operate continuously.
- Projects announced in the same industrial corridor may compete for the same network capacity.
Malaysia’s renewable-electricity composition targets are 31% by 2025, 40% by 2035 and 70% by 2050. Meeting those targets while serving round-the-clock computing demand requires both new clean generation and efficient facilities; buying renewable certificates alone does not remove the need for physical capacity at the connection point.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How can data centres reduce cooling-water use?
SPAN’s guidelines encourage a hierarchy that starts with demand reduction and ends with reuse. Developers and approving agencies can apply the following checks during design and operation.
Choose cooling for the local climate and workload
High-efficiency cooling, including systems that match cooling output to actual computing load, reduces both electricity overhead and evaporative losses. A design should publish expected water-use effectiveness for normal, hot-weather and peak-load conditions rather than quoting one laboratory value.
Use non-potable sources where safe
Rainwater harvesting, treated wastewater and other reusable-water applications can displace treated drinking water for cooling-tower make-up and maintenance. Source quality, storage capacity and public-health controls must be specified so that substitution does not create a different reliability problem.
Recycle cooling-tower blowdown
The Cycle of Concentration approach allows operators to run cooling water through more cycles before blowdown, subject to corrosion, scaling and biological controls. Higher cycles can reduce fresh-water demand, but only when treatment and monitoring maintain equipment safety.
Set operating conditions, not just construction promises
Water meters should separate potable make-up, recycled-water use, discharge and evaporation. Monthly reporting can then show whether a facility meets its approved water-use limit during peak computing periods. Approval conditions should include corrective actions when consumption exceeds the design envelope.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →A practical way to compare data-centre proposals
Public agencies, utilities and communities can evaluate proposals on the same six dimensions instead of relying on the headline investment value.
| Decision axis | Questions to ask |
|---|---|
| Electricity load | What is the maximum demand, annual energy use and peak-demand timing? How quickly will each phase be energised? |
| Efficiency and renewables | What PUE is guaranteed under local weather conditions, and what renewable supply is physically or contractually available? |
| Cooling system | Is cooling air-based, evaporative or hybrid? What is the expected water-use effectiveness at full load? |
| Water source | How much demand is potable, and how much can come from recycled, rain or seawater sources? |
| Local capacity | What are the current reserve margin, treatment-plant capacity, pipeline limits and substation constraints? |
| Cost allocation | Which grid and water upgrades are paid by the developer, and which remain regulated public assets? |
What to watch through 2035
The most useful indicators are actual connections rather than announcements: TNB’s energised load, PETRA’s revised peak-demand forecasts, water allocations that become operational, and measured cooling-water intensity. Johor’s treatment-plant and pipeline schedules deserve particular attention because a small reserve margin leaves less tolerance for delays or drought.
Data-centre growth can support Malaysia’s digital economy without destabilising basic services if approvals are phased, infrastructure costs follow the users who create them, and every facility meets enforceable electricity and water-efficiency conditions. The outcome will be determined less by the number of buildings announced than by where they connect, how they cool, and whether supporting capacity arrives first.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
