A consortium of Gamuda, SD Guthrie and Gentari signed a term sheet on 28 September for a 21-year bilateral energy supply contract with a United States-based multinational technology company, covering a hybrid solar plant of at least 680 MWac in Perak paired with a four-hour battery system. Commercial operation is targeted for 2029. The companies estimate the contract will generate more than RM10 billion in gross revenue for the project company over its term.
The offtaker was not named, but the procurement route tells most of the story: the electricity will be sold to that company's data centers in Malaysia under the Corporate Renewable Energy Supply Scheme, or CRESS. Malaysia has spent two years building that scheme into a working mechanism, and the Perak deal is the clearest test yet of whether it can carry hyperscale load.
The contract sits with G3nerasi Kinta Sdn Bhd, wholly owned by G3nerasi Mutiara Sdn Bhd — a joint venture incorporated on 8 September 2026 in which SD Guthrie Renewable Energy holds 33.5%, Gentari Renewables 33.5% and Gamuda Energy 33%. Funding is expected to come from project financing plus equity contributions from the three partners in proportion to their stakes. Execution of the supply contract and the remaining CRESS agreements is targeted for the first quarter of 2027, and the 2029 commercial operation date remains conditional on a renewable energy supply access agreement with grid owner Tenaga Nasional and the required CRESS approvals.
The partnership has been forming for a while. Gamuda and SD Guthrie agreed in August 2025 to develop solar assets with up to 1.2 GW of accumulated capacity; Gamuda struck a separate partnership with Gentari the same month for 1.5 GW. The division of labor is visible in the ownership: SD Guthrie brings land, Gentari brings renewables development, Gamuda brings construction.
CRESS was introduced in 2024 as a third-party access mechanism. Rather than putting solar on a customer's roof, it lets a renewable developer generate anywhere in Peninsular Malaysia and deliver through the TNB grid to a corporate buyer, paying a System Access Charge for the use of the network. The charge has been the scheme's moving part.
| Policy step | System Access Charge (firm / non-firm) | Other conditions |
|---|---|---|
| CRESS launched, 2024 | 25 / 45 sen per kWh | New grid users only |
| Revised, 29 August 2025 | 20 / 40 sen per kWh | Existing commercial and industrial customers admitted from March 2025 |
| CRESS Acceleration Package, September 2026 | 14 sen per kWh (firm) | Minimum 10-year contracts; commercial operation required by 31 December 2028 to qualify, with no extensions considered |
Two technical conditions matter as much as the price, and together with the two-rate structure they are the whole mechanism. To be classified as firm output, a solar project must pair with battery capacity of at least 50% of its export capacity, capable of sustaining output for four consecutive hours. The gap between the rates is what makes the rule bite: non-firm supply has consistently been priced at close to twice the firm rate, and now sits close to three times as high against the 14 sen package. A project that cannot deliver firm output pays that penalty for the same electrons, and a project whose battery is unavailable or falls short of the specification drops to the non-firm rate. Delivery is still time-limited: the 14 sen rate applies only to projects reaching commercial operation by 31 December 2028, and the Energy Commission is directed to refine the technical requirements covering solar and battery configuration, system operation and access to grid connection data.
Why four hours, and not two or six, is not explained in the published rules. The design intent is visible in how the charge was first framed in 2024, when the 25 sen rate was offered for output where the grid owner did not have to deal with intermittency and received energy much as it would from a conventional plant, against 45 sen for intermittent supply. The requirement was left to the Energy Commission as an administrative test rather than a system study. The four-hour figure lines up with the window in which a plant has to cover for itself once solar falls away — Peninsular Malaysia’s time-of-use peak runs to 22:00, and solar output is effectively gone by about 18:00 — but we have found no published derivation, and the Commission’s own guidance still poses how long firm capacity must be sustained when it is declared a day ahead as an open question. Read it as the current threshold for a rate, not a settled engineering constant.
Take-up so far is registered, not built: 11 renewable developers and eight green consumers had signed up with the Single Buyer as of September 2026, representing about 3,148 MW of project capacity. Analysts estimate that pipeline implies roughly RM16 billion of engineering, procurement and construction opportunity at about RM5 million per MW.
Malaysia's storage case is being built by server racks. Data centers accounted for about 9.28% of national electricity consumption as of mid-August 2026, up from roughly 6% in the first half of the year. TNB had secured 61 data center projects with a combined maximum demand of 8.35 GW by June 2026; 42 projects totalling 5.65 GW were connected to the grid, but actual load was running at only about 1.26 GW because campuses were still filling with equipment. A further 16 projects totalling 2.21 GW were under construction, and three totalling 0.49 GW had signed supply agreements. Johor accounts for about 67% of secured capacity. Government projections put data center demand near 21 GW by 2040.
That load profile is what makes storage non-negotiable. Analysts cite TNB guidance that once solar penetration passes 6 GW, each additional gigawatt of solar requires about 0.5 GW of battery capacity to keep the system stable. Meanwhile, newly approved data center projects are reported to face a minimum renewable energy requirement of about 30%, a figure that comes from analyst channel checks rather than a published rule and should be treated as indicative.
On the Perak project, the firm-output rule sets a floor on the battery: at 680 MWac of export capacity, the minimum compliant battery would be about 340 MW / 1,360 MWh. The consortium has not disclosed the actual battery size, and in practice a 21-year hyperscale contract of this kind is unlikely to be sized to the regulatory minimum.
Analyst estimates put equity internal rates of return for firm-supply CRESS projects in the 10-13% range, assuming a developer tariff near 35 sen per kWh against the 14 sen access charge. That spread is thin enough that the trajectory of the access charge over a 21-year contract becomes a financing variable in its own right. The charge is fixed for the current regulatory period, 2025 to 2027, with future variation previously capped at 15% per period — but at least one research house has flagged that the September package did not explicitly confirm the three-year review and the 15% cap will continue, leaving escalation risk-sharing as a live negotiation point.
There is also a queue problem behind the headline. The 2028 deadline applies to projects qualifying for the reduced charge, and Malaysian developers are already working through Large Scale Solar rounds LSS5, LSS5+ and the forthcoming LSS6. Analysts expect the compressed timetable to accelerate CRESS negotiations, with engineering contracts possibly landing from the fourth quarter of 2026 onward — which puts pressure on equipment delivery slots for projects that must be built and energised within roughly 21 months of the intended first-quarter 2027 financial close.
The 680 MWac minimum capacity, the four-hour battery specification, the 21-year term, the RM10 billion revenue estimate, the joint venture ownership and the 2029 commercial operation target are confirmed in the listed exchange filings as reported by at least three independent outlets. The firm-output rule of 50% of export capacity with four-hour endurance is the Energy Commission’s own standing requirement, cross-checked against trade coverage, and the minimum compliant battery size of about 340 MW / 1,360 MWh is our calculation from that rule applied to disclosed project capacity, not a disclosed figure. Three items are weaker than the rest: the roughly 30% renewable requirement for newly approved data centers comes from analyst channel checks rather than a published regulation; the roughly RM16 billion EPCC estimate is an analyst figure; and our reading of the four-hour threshold as aligned to the post-solar evening window rests on the published time-of-use structure, not on any published derivation of the threshold itself. The future treatment of the system access charge beyond the current regulatory period is unresolved in the public record.