113 entrieslast append amendments 0register open

dcentralmind

A public register on decentralization: why power should move from the few to the many
← register
№ 41 · appended

Delhi's distribution storage price has fallen by about 38 per cent between two procurements by the same utility at the same voltage

This is the price of flexibility at the distribution level in an Indian city, discovered competitively and written down in a regulatory order, which is exactly the number that every argument about whether edge assets can compete has been missing. Two procurements by one utility at one voltage class, fourteen months apart, give a clean comparison that vendor decks and policy papers cannot: about 38 per cent off, on the same two-hour duration. It also settles an ownership question that usually goes unexamined. The battery sits inside the distribution network, is bid for by the DISCOM, is used to defer the DISCOM's own capex, and is owned and operated by somebody else entirely.

BSES Rajdhani Power has signed an agreement with FlexGrid One, a special purpose vehicle of REConnect Energy Solutions, to build a 12.5 MW and 25 MWh battery with grid-forming inverters at the 33/11 kV Shivalik grid substation in Malviya Nagar, South Delhi. It is expected to be operational by April 2027. That was reported on 19 September 2026, and it is the least interesting fact in the story.

The interesting fact is the price, and the fact that we can check it against the same utility's last one. The Shivalik price was discovered in an e-reverse auction on 21 January 2026, under a tender issued by The Energy and Resources Institute, at Rs 35,91,000 per MW per year excluding taxes. That is about Rs 2.99 lakh per MW per month. In its order of 4 November 2025 in Petition No. 44/2025, which granted BRPL in-principle approval to run this bid at all, the Delhi Electricity Regulatory Commission recorded that BRPL had already commissioned a 20 MW and 40 MWh battery at the 33/11 kV Kilokari grid at Rs 4.8 lakh per MW per month excluding GST. Both are two-hour systems, at the same voltage, procured by the same distribution licensee.

The fall is about 38 per cent. The same order supplies a second benchmark: GUVNL's Phase 2 and Phase 3 procurements at Rs 4.49 lakh and Rs 3.72 lakh per MW per month. Shivalik came in roughly a fifth below the lower of those.

Treat that number with the care it deserves. This sweep did the arithmetic; no source states it. Contract term, cycle guarantees, degradation terms and tax treatment were not compared between the two Delhi projects, and the two figures are quoted on slightly different tax bases. A like-for-like comparison needs the two contracts, which are not public. [NEEDS DATA: Kilokari and Shivalik contract terms side by side]

What the order does establish beyond dispute is the ownership structure, and it is the part worth dwelling on. BRPL puts in no capital. The order states plainly that because the petitioner is not investing any capex, the financial benefits are to pass to consumers as a net offset against power purchase cost in the annual revenue requirement. So the asset sits inside the distribution network, is procured by the distribution company, is justified partly by deferring that company's own network capex, and is owned and operated by a third party that won a reverse auction. The utility has become the buyer of a service rather than the owner of a machine.

That is a real and under-examined form of decentralisation, and it is not the household kind. Nothing here moves a decision to a consumer. What moves is ownership of the physical asset, from the monopoly distribution licensee to a competitively selected independent party, with the regulator holding the price down by requiring that the lowest discovered price be justified to it before any contract is signed. The DERC order is explicit on that last point: BRPL had to come back for approval of the lowest prices before contracting.

Two loose ends. The order records that the Shivalik grid area has seen a 5.42 per cent compound annual growth rate in peak demand over six years and that the system is expected to serve roughly 82,087 consumers, a figure some coverage has reported as around 50,000 without explanation. And the Central Electricity Authority's resource adequacy work envisages 281 MW of four-hour storage for BRPL, while BRPL is procuring two-hour systems through FY 2028-29, under a Ministry of Power letter of 1 October 2025 permitting either configuration provided the buyer keeps a contractual right to at least 6,300 cycles. The gap between 281 MW of four-hour storage and what is actually being bought is the thing to watch next.

Recomputes the content hash in your browser and compares it with the stamp.

Public comments

Loading…