Italy Raised Its Battery De-Rating Factors and Cut Them. Both Headlines Are Correct.

On 19 October 2026, one trade publication reported that Italy had raised battery de-rating factors for the 2028 capacity market. On 20 October, the same publication reported that Italy had cut capacity market ratings for four-hour batteries in the south. Both headlines describe the same ministerial decree, and both are correct. The word in the middle — “de-rating factor” — is doing opposite work in each sentence, and that ambiguity is not confined to Italy. It sits inside the accreditation rules of almost every capacity market a storage developer can bid into.

What Decree 327 actually changed

Italy replaced a single national de-rating value for batteries with a matrix that varies by market zone and by energy-to-power ratio, and the new numbers are materially worse for new projects than the version approved two months earlier. Decree No. 327, dated 16 October 2026, was issued by Italy’s Ministry of Environment and Energy Security (MASE) and adopts a proposal that transmission system operator Terna submitted on 18 September 2026. The energy regulator ARERA gave a positive opinion in Resolution 370/2026/R/eel on 14 October 2026, limiting its approval to the 2028 delivery auction. The ministry’s stated reason: Terna’s updated data showed a significant increase in authorised or pending-authorisation storage capacity compared with the figures used in its 25 June 2026 proposal.

The decree also set the import limit from the northern foreign virtual area — covering France, Switzerland, Austria and Slovenia — at 1,700 MW, with 113 MW for the south-central area (Montenegro), 52 MW for the southern area (Greece) and zero for the remainder. National probabilistic available-capacity requirements run from 66.2 GW at point B to 67.8 GW at point D, of which roughly 39.7 GW sits in the north.

Why two outlets reported opposite verbs on the same document

Because the published figures for batteries and the published figures for other technologies are not on the same basis, and nothing in the headline tells you which basis you are reading. For a four-hour battery, the reported number is the recognised capacity — the share of nominal power that counts as firm in the auction. For new solar, the reported figure is 91.7%; for wind 84%; for run-of-river hydro 67%; for new thermal plants 10%, rising to 20% for combined-cycle plants cooled exclusively with water.

Those two sets cannot mean the same thing. If 91.7% were recognised capacity, new Italian solar would be firmer than a four-hour battery in every zone in the country and nearly as firm as nuclear is rated in PJM — which is not a defensible reliability claim for a plant that produces nothing after sunset. The solar, wind, hydro and thermal figures are the share removed; the battery figures are the share kept. An Italian trade report on the ARERA opinion put the mechanics plainly in its own headline: the coefficient goes up, the remuneration goes down. So “Italy raised de-rating factors” and “Italy cut four-hour ratings in the south” are the same sentence read from opposite ends.

The zonal table, and why the far south lost the most

A four-hour battery now accredits at 21% of nominal power in the south, Calabria, Sicily and Sardinia — against 67% nationally in the 2027 auction. The zone ranking also inverted: the south-central zone, at 51%, now accredits more than the north, at 40%.

Zone	4h: 2027 auction	4h: Aug 2026 version	4h: Decree 327 (16 Oct 2026)	8h: Aug 2026 version	8h: Decree 327	Source
North	67% (national)	51%	40%	not reported	not reported	MASE Decree 327, reported by pv magazine Italia and ESS News, 19–20 Oct 2026; August values per Aurora Energy Research analysis
Centre-North	67% (national)	54%	40%	not reported	not reported	As above, 19–20 Oct 2026
Centre-South	67% (national)	not separately reported	51%	73%	71%	As above, 19–20 Oct 2026
South and Calabria	67% (national)	40%	21%	52%	29%	As above, 19–20 Oct 2026
Sicily and Sardinia	67% (national)	31%	21%	47%	33%	As above, 19–20 Oct 2026

Two cautions on this table. The August-version figures were reported as a sequence of four values — 51%, 54%, 40% and 31% — and only the north and the south mappings were stated explicitly; the middle two are the order in which the zones were listed, not an independently confirmed pairing. And eight-hour values for the north and centre-north were not published in either report. Where a cell says “not reported,” it means exactly that: not that the figure does not exist.

Doubling the duration does not double the accredited megawatts

Moving from four hours to eight hours raises accredited power by roughly 38% in the south and 39% in the centre-south, and by about 57% in Sicily and Sardinia — for twice the installed energy. Those are our calculations from the decree’s published percentages (29% versus 21%, 71% versus 51%, 33% versus 21%), not ratios stated in the decree.

The practical consequence is that duration buys capacity-market revenue at very different exchange rates depending on where the asset sits, and in no zone does it buy it proportionally. A developer sizing a project on a national rule of thumb — “longer duration earns a better factor” — gets the direction right and the magnitude wrong by a factor that varies across the country. This is the same failure mode we have flagged in auction headlines elsewhere: a single national number travels further than it should, as it did in Brazil’s first battery storage auction.

The clause that gives two accredited capacities to the same battery

The new coefficients do not apply to everyone, and the carve-out is the single most consequential sentence in the decree. The revised factors apply only to storage capacity that was absent from the 2027 auction supply curve, or that was offered and not selected. Existing capacity, capacity already contracted in previous auctions, and lithium-ion batteries awarded in the MACSE storage procurement auction for 2028 delivery all retain the 2027 auction de-rating factors.

So two four-hour batteries of identical specification, in the same Italian zone, can carry accredited capacities of 67% and 40% of nominal power — a 27-percentage-point gap created entirely by paper history. For anyone valuing an Italian storage portfolio, the de-rating factor is now an asset-level attribute that has to be traced through auction records, not a zone-level lookup. Grandfathering clauses of this kind are routinely omitted from secondary coverage; they are also where the money is, which is the same lesson as the eligibility conditions buried in the U.S. storage tax credit rules.

Three markets, one word, three different things measured

Capacity accreditation is published under at least three incompatible conventions across major markets, so a cross-market comparison built from headline percentages is arithmetic on mismatched units.

Market	Published term	What the number measures	Reference value	Date / document
Italy (2028 delivery)	De-rating factor / recognised capacity	Batteries published as share kept; solar, wind, hydro, thermal published as share removed	4h battery 21%–51%; new solar 91.7% removed	MASE Decree 327, 16 Oct 2026; ARERA Resolution 370/2026/R/eel, 14 Oct 2026
PJM (2026/2027 delivery)	ELCC class rating	Share of nameplate accredited, by resource class	4h storage 50%; 8h 62%; 10h 72%; fixed-tilt solar 8%; gas CC 74%; nuclear 95%	PJM, ELCC Class Ratings for the 2026/2027 Base Residual Auction
Great Britain	De-rating factor	Share of nameplate recognised, by storage duration	1h battery 8.4% (2028), 10.3% (2029) per Modo Energy’s published methodology documentation; CCGT around 91%	Analyst figures, accessed Oct 2026; NESO publishes the official table in auction guidelines
Spain (new market)	Firmness coefficient	Coefficient between 0 and 1 applied to installed capacity to derive firm capacity; only technologies with a defined coefficient may bid	Values not yet published	Ministerial Order TED/966/2026, 15 Sept 2026; BOE 17 Sept 2026

Note what Spain tells you: the coefficient framework exists in force, with contracts of up to 15 years for new investment, a 1 MW minimum firm-capacity bid block and escalating penalties for failed availability tests — and the numbers that decide every bid are still to come. A market can be fully legislated and completely unmodellable at the same time.

What this does to the bid arithmetic

At an unchanged clearing price, a 100 MW / 400 MWh battery in northern Italy sees accredited capacity fall from 67 MW under the 2027 factors to 40 MW — a 40% reduction in capacity-market revenue per megawatt installed. In the south, Calabria, Sicily or Sardinia the same asset falls to 21 MW, a 69% reduction. Those percentages are our arithmetic on the published coefficients; they assume nothing about how the 2028 auction clears.

That matters because the clearing price is the one variable in the stack that the de-rating change itself will move. Lower accreditation per unit of nameplate means more nameplate has to be built, or cleared, to fill the same probabilistic requirement — and the decree pairs the new coefficients with a rule letting successful new-capacity bidders postpone the start of their delivery period to December of the first delivery year, on notice to Terna by 30 April of the preceding year, with an adjustment auction to follow within 60 days. Volume, timing and accreditation all moved in the same document. Anyone carrying a pre-October Italian capacity-revenue model forward is now running it on three stale inputs, not one. The framework for rebuilding that stack is the one we set out in our BESS market sizing framework.

How to check an accreditation factor before you model it

Four questions settle whether a published percentage means what you think it means, and all four can be answered from primary documents. First: is the figure the share kept or the share removed? Compare the technology set — if solar outranks four-hour storage, you are reading removals. Second: what is the unit of variation — national, zonal, duration, resource-specific, or some combination? Italy now varies by zone and duration; PJM varies by class with resource-specific values for some gas units. Third: which vintages does it bind? Ask explicitly whether existing, contracted and separately procured capacity is grandfathered, and at which prior factor. Fourth: which delivery year does it apply to, and is the approval limited to that year? ARERA’s October opinion was expressly limited to the 2028 auction, which means the 2029 numbers are a separate fight.

The primary record for the Italian figures sits with the regulator: ARERA publishes its resolutions, including 370/2026/R/eel, in its official register of acts and measures. We have not read the decree text itself, only the two trade reports of it and the regulator’s opinion as those reports describe it; the zonal percentages above are consistent across both reports, and the sign convention is an inference from the arithmetic, not a quotation from the decree. If you are committing capital, read Annex 4 of the decree yourself.

Figures verified 21 October 2026. Policy changes fast — check the primary source.