Hardware stopped being the obstacle
BloombergNEF's 2021 Lithium-Ion Battery Price Survey, published 30 November 2021, put the volume-weighted average pack price at USD 132 per kWh in real 2021 dollars, down 6% from USD 140 per kWh in 2020. Packs for battery electric vehicles came in at USD 118 per kWh and cells at USD 97 per kWh, with cells accounting for roughly 82% of pack cost. Against the same survey's starting point of above USD 1,200 per kWh in 2010, that is an 89% decline in real terms.
For a Mexican industrial buyer the useful reading of that series is not the headline decline. It is the 6%. After a decade of double-digit annual falls, the 2021 survey recorded the slowest drop BNEF had measured, and BNEF attributed the slowdown to rising commodity prices feeding into cell costs. Anyone building a capital plan on the assumption that waiting two years halves the price of a battery is extrapolating a curve that has visibly bent.
That matters because the constraint on storage in Mexico has moved. Equipment cost is no longer what stops a project. What stops it is that Mexican regulation gives a battery almost nothing to sell, and until the Comisión Reguladora de Energía changes that, the only storage projects worth underwriting here are the ones whose entire return sits behind the customer's own meter.
The cost curve, with its basis stated
Battery price figures are quoted carelessly, and the difference between the numbers is large enough to change an investment decision. Three distinctions are worth holding. A cell price is not a pack price, and the gap in the 2021 survey was USD 97 against USD 118 per kWh for electric vehicle applications. A pack price is not an installed system price for a stationary project, which adds inverters, containers, thermal management, protection, civil works and interconnection, and which BNEF's vehicle-weighted average does not attempt to cover. And a nominal price is not a real one, which is why the survey states its figures in real 2021 dollars.
The practical consequence for a plant in Mexico is that the delivered cost of a behind-the-meter system is set mostly by items that are not falling at battery-cell speed. Switchgear, transformers, engineering and installation labor are local costs, and in 2021 and 2022 they were rising. We would expect a project evaluated today on a cell price rather than an installed cost per usable kWh to come in materially over budget, and the error compounds where the site also needs a service modification from CFE Distribución.
What Mexico's auctions proved, and what they stopped proving in 2018
Mexico ran three long-term auctions under the market design established by the 2014 reform, and their results are the reason the country is still described as a natural home for renewable generation.
| Auction | Results announced | Capacity awarded |
|---|---|---|
| SLP-1/2015 | March 2016 | 1,720 MW |
| SLP-1/2016 | September 2016 | approx. 2,871 MW |
| SLP-1/2017 | 22 November 2017 | 2,562 MW across 14 projects |
The third auction awarded 1,323 MW of solar photovoltaic capacity across nine plants, 689 MW of wind across five, and 550 MW of natural gas at one. Its most economical package cleared at an average of USD 20.57 per MWh bundled with one Certificado de Energía Limpia. That bundling matters. The bid buys energy and a CEL together, so quoting the figure as a bare energy price overstates how cheap the electricity alone was.
The record low bid in that auction, roughly USD 17.7 per MWh plus one CEL, is attributed to Enel in all contemporaneous reporting, and the sources disagree on the technology. Reuters Events reported it in November 2017 as a world-record low solar price. Energía Estratégica reported USD 17.77 per MWh for wind, secured by Enel Rinnovabile for the Energía Limpia de Amistad 2 and 3 projects in Coahuila. We do not assert either attribution here, because CENACE's official fallo document for SLP-1/2017 with the per-project price and technology table is what would settle it, and we have not been able to retrieve it. The price and the auction are well attested. The technology is not.
The more consequential fact is what happened next. CENACE suspended the fourth long-term auction on 3 December 2018 and announced its cancellation on 1 February 2019. No further long-term auction has been held. Whatever storage is built in Mexico will not be built on the back of an auction-awarded pipeline, which removes the mechanism that made the first wave of renewable projects financeable.
The statutory target is clean energy, not renewable energy
Mexico's electricity generation target is set by the Ley de Transición Energética, whose Transitorio Tercero requires a minimum 35% clean-energy share of electricity generation by 2024, with interim steps of 25% by 2018 and 30% by 2021. The word matters. Energías limpias, as defined in article 3 of the Ley de la Industria Eléctrica, is broader than renewables. It includes nuclear generation, efficient cogeneration meeting CRE efficiency criteria, thermal plants with geological carbon capture and storage, and several waste and hydrogen pathways.
Two things follow for a buyer. Laguna Verde's output counts toward the target, so progress against 35% does not measure wind and solar build-out. And an efficient cogeneration plant on an industrial site can qualify as clean generation under the same definition, which makes it a live alternative to a solar-plus-storage configuration for facilities with a steady thermal load. Anyone comparing decarbonization routes in Mexico should price both.
What CRE would have to issue for storage to have a market
Naming the instrument is more useful than predicting how many will appear. For a battery at an industrial site or on the grid to earn market revenue in Mexico, CRE would have to issue disposiciones administrativas de carácter general that define an electricity storage system, assign it a permit class, and set how energy drawn to charge and energy discharged are settled. CENACE would then have to create a corresponding market product, either in the mercado de balance or in servicios conexos, so that a battery has something to bid into and a settled price to be paid at.
Neither existed in mid-2022. Until both do, several revenue lines that appear routinely in storage proposals have no legal basis in Mexico. Payment for frequency response, capacity payments to a standalone battery, and arbitrage between the Mercado del Día en Adelantado and the Mercado de Tiempo Real all presuppose a market role that Mexican regulation has not yet created. A finance director reviewing a storage proposal should ask which published CRE instrument creates each revenue line, and delete any line where the answer is a forecast rather than a document.
Where the value is available now
Strip out the unavailable revenue and a real case usually remains. Behind the meter, a battery can cut the monthly capacity charge on a GDMTH, DIST or DIT account by shaving the site's own demand peak, can hold critical process load through an interruption, and can ride through the voltage events that trip drives and damage motors. None of it requires a market rule. All of it is measurable from data the facility already has.
Paired with an onsite array, storage also changes what generation is worth. Under medición neta, surplus solar accrues as credits against future consumption for up to twelve months under CRE Resolución RES/142/2017 (DOF, 7 March 2017), which reduces energy charges but does nothing for a capacity charge set after sunset. Storage is what converts an afternoon surplus into an evening peak reduction. That is the specific mechanism by which storage adds value to a renewable installation in Mexico, and it is worth modeling before the array is sized rather than after.
The policy and lithium side of this question, including what the April 2022 Ley Minera reform did and did not change, we take up separately in our note on lithium policy and the storage regulatory gap. Degradation remains the open technical variable. Warranted cycle life and throughput guarantees vary widely between suppliers, and the warranty terms, not the datasheet, are what a lender will read.
Size a storage system against your own load, not a market forecast
Mexico Energy Partners can model onsite solar with energy storage for a specific facility, sizing the battery against measured demand peaks and stating the installed cost per usable kWh rather than a cell price. The initial review needs twelve months of CFE billing with demand by period, interval data at the main meter, a single-line diagram and a list of loads that cannot tolerate an interruption. Results vary with tariff class, region, load factor and consumption profile, and no savings range can be confirmed before that data is reviewed.