A 300MWh liquid air “battery” in Manchester shows how long-duration energy storage and AI-backed optimization can anchor a resilient, low-carbon power system.
Why a 300MWh “air battery” in Manchester matters
Greater Manchester is building something that doesn’t look like a battery at all—and that’s exactly the point.
Highview Power has just broken ground on a 50MW / 300MWh liquid air energy storage (LAES) plant at Carrington, billed as the world’s largest commercial facility of its kind. It doesn’t sit in shiny racks like lithium-ion cells. It looks more like industrial gas equipment. Yet it could do something the UK grid badly needs: store clean energy for hours at a time, reliably, for decades.
For anyone serious about green technology, this project is a big signal. Short-duration batteries helped solar and wind scale. Long-duration energy storage (LDES) like liquid air will decide how far we can push fossil fuels off the system.
This article breaks down what Highview is building, how liquid air energy storage actually works, where AI fits into running assets like this, and what this shift means for utilities, cities, and energy-intensive businesses.
What Highview is building in Carrington
Highview’s Carrington project is a 50MW / 300MWh LAES plant located at Trafford Low Carbon Energy Park near Manchester, scheduled to start operating in late 2026.
Key facts at a glance
- Capacity: 50MW power, 300MWh energy (around 6 hours at full output)
- Location: Carrington, Greater Manchester, UK
- Technology: Liquid air energy storage (LAES)
- Lifetime: Designed for 40–50 years with no performance degradation
- Role: Long-duration, dispatchable, zero-emission power and grid stability
The plant connects into existing substation and transmission infrastructure, which keeps project costs and local disruption down. Highview is also building what it calls a “stability island”—essentially a set of synchronous condensers and power electronics that provide inertia, voltage support, and fault level to the grid.
That stability piece might sound like a technical footnote, but it’s a big deal. As coal and gas plants retire, grids are losing the physical spinning mass that keeps frequency stable. Projects that combine long-duration storage + grid stability services are becoming some of the most valuable green infrastructure assets on the system.
Funding-wise, Highview has attracted heavy hitters: £300 million from the UK’s National Wealth Fund, Centrica, Rio Tinto, Goldman Sachs, KIRKBI and others for Carrington, plus £130 million for a similar stability-focused project in Scotland.
The message from investors is clear: this isn’t a science project anymore. LAES is moving into the commercial asset class.
How liquid air energy storage actually works
Liquid air energy storage is mechanical energy storage that uses air instead of lithium, vanadium, or exotic chemistries.
Step-by-step: turning air into a long-duration battery
At a high level, LAES cycles through three phases:
-
Charging (store energy)
- When there’s excess renewable energy—e.g., strong wind overnight—electricity drives industrial chillers.
- These chillers cool ambient air down to around –196°C, turning it into a liquid.
- The liquid air is stored in insulated, low-pressure tanks.
-
Storage (hold for hours, days, or weeks)
- Liquid air sits in tanks much like liquefied natural gas, but without the carbon.
- Storage duration is flexible: hours, days, potentially weeks, limited mainly by tank size and system economics.
-
Discharging (generate power)
- When power is needed, the liquid air is pumped to high pressure and warmed.
- As it warms, it expands rapidly—up to ~700 times its liquid volume.
- This expansion drives a turbine to generate electricity, just like a conventional power plant but without fuel combustion.
The round-trip efficiency of LAES is lower than top-tier lithium-ion (typically in the 50–65% range, depending on design and heat integration), but that’s only half the story. LAES brings three strategic advantages that matter at grid scale:
- No degradation over 40–50 years of operation
- No critical metals or flammable electrolytes
- Flexible duration without redesigning the core technology—tanks and auxiliaries can simply be scaled
For national grids wrestling with seasonal swings in wind and solar, that matters more than a few percentage points of efficiency.
Why long-duration storage is the missing piece for renewables
Here’s the thing about renewable integration: adding more wind and solar without adding long-duration storage creates diminishing returns.
Once a grid reaches a certain level of variable renewables, three problems show up:
- Curtailment: Wind and solar farms are told to switch off because the system can’t absorb more power. In some markets, curtailment can hit 5–20% of potential output.
- Capacity gaps: On calm winter evenings, you’ve got high demand, low solar, and often low wind. The result: gas peaker plants, imports, or blackouts.
- System stability: Retiring fossil plants means losing inertia and voltage support. Even with enough energy overall, the grid can still become unstable.
Projects like Highview Carrington are designed to tackle all three:
- Store excess renewables instead of curtailing them, and release them when wholesale prices spike.
- Provide firm capacity over 4–8 hours, replacing or deferring gas peakers.
- Deliver grid-forming and stability services via the stability island.
For the UK specifically, this aligns with the ongoing shift away from unabated gas and the need to back up major offshore wind build-out. In plain language: if we want a renewables-heavy grid that still works on a cold, windless January evening, we need long-duration energy storage projects on the ground—not just in PowerPoint decks.
Where AI fits: making LAES smarter and more profitable
You can build a great long-duration storage asset and still run it poorly. That’s where AI and advanced analytics quietly become the difference between a marginal project and a strong one.
1. Dispatch optimization
LAES plants make money by buying electricity when it’s cheap (or even negatively priced) and selling when it’s expensive, plus providing grid services. That’s a classic optimization problem where AI excels.
A well-designed AI-based dispatch engine can:
- Forecast wholesale prices, renewable output, and demand curves
- Decide when to charge, when to discharge, and when to sit idle
- Co-optimize multiple revenue streams (energy, capacity, ancillary services)
In practice, this often increases revenue and asset utilization while reducing wear on auxiliary systems like pumps and compressors.
2. Predictive maintenance
Mechanical systems—compressors, cryogenic tanks, turbines—are robust but not immortal. AI-driven predictive maintenance uses sensor data to:
- Detect early signs of mechanical or thermal issues
- Predict failure probabilities for critical components
- Recommend optimal maintenance windows that avoid high-revenue hours
For a 40–50 year asset, even a 2–3% improvement in availability translates into major financial gains over the lifetime.
3. Portfolio and system planning
As more green technology assets like LAES, large-scale batteries, and flexible demand come online, planning becomes a multi-variable puzzle. AI-based planning tools help utilities, cities, and large energy users answer concrete questions:
- How many hours of storage do we actually need in our region? 4, 8, 12+?
- Where should assets like Carrington be sited to unlock the most grid value?
- What mix of storage technologies—lithium-ion, LAES, pumped hydro, CAES—delivers resilience at the lowest system cost?
I’ve found that teams who bring AI into planning early are less likely to overbuild the wrong technology and more likely to de-risk investments.
What this means for utilities, cities, and large energy users
The Highview Carrington project isn’t just a UK headline; it’s a template. If you’re responsible for energy strategy anywhere, there are some practical lessons here.
For utilities and grid operators
Utilities can treat LAES and other long-duration assets as flexible infrastructure, not just “big batteries.” That means:
- Designing tenders that explicitly value 4–8+ hour duration, not just MW
- Procuring stability and inertia services alongside energy capacity
- Using AI tools to simulate how LAES changes congestion and curtailment patterns
The utilities that move early on long-duration storage tend to gain negotiating power and regulatory goodwill because they’re solving real system problems instead of simply adding megawatts.
For cities and regions
Greater Manchester is positioning Carrington as part of a low-carbon energy hub, which is a smart move for any region with net-zero goals.
City and regional planners can:
- Integrate long-duration storage into heat networks, hydrogen hubs, and industrial clusters
- Use these assets to support data centres, electrified transport depots, and critical services
- Align planning rules, zoning, and grid upgrades around multi-decade assets like LAES
Done right, projects like Carrington become anchor points for green jobs and clean industry, not just technical curiosities.
For large energy users and corporates
If your business consumes a lot of power—manufacturing, logistics, data, chemicals—long-duration storage unlocks options beyond standard PPAs.
You can:
- Structure long-term offtake agreements that tie your demand to storage-backed renewables
- Improve resilience by co-locating with or contracting capacity from assets like LAES plants
- Use AI-driven energy management systems to align flexible loads with storage dispatch
Companies that pair strong sustainability targets with smart contract structures around assets like Carrington typically get both lower long-term costs and a more credible decarbonization story.
Where liquid air fits among other non-lithium options
LAES is part of a wider shift toward non-lithium energy storage:
- Pumped hydro: Very mature, high efficiency, but location-constrained and slow to build.
- Compressed air energy storage (CAES): Uses underground caverns; promising but geology-dependent.
- Flow batteries: Great for 4–10 hour durations with very long cycle life; chemistry still evolving.
- Thermal storage: Ideal where heat and power can be co-optimized (district heating, industry).
Liquid air’s niche is clear:
- Works well where industrial skills and gas-handling expertise already exist
- Scales like infrastructure, not consumer electronics
- Fits into brownfield industrial sites, ports, and energy parks
Is it the answer everywhere? No. But in markets like the UK, parts of Europe, and dense industrial regions worldwide, LAES is likely to become a key piece of the long-duration mix.
The bigger story: green technology is becoming infrastructure
The reality is simpler than it looks: we don’t get a reliable net-zero grid without long-duration storage. The Highview Carrington project is one of the first serious attempts to build that backbone using liquid air energy storage at scale.
From a green technology perspective, it checks several boxes:
- Uses abundant air instead of scarce minerals
- Supports deeper renewable integration and lower curtailment
- Delivers grid stability services usually tied to fossil plants
- Runs for decades, making it a true infrastructure asset
If you’re planning your own energy transition roadmap—whether as a utility, city, or corporate—the next step is straightforward: start modeling what happens to your system when you add 4–12 hours of long-duration storage backed by AI-driven optimization. Use that to challenge your current assumptions about peaker plants, imports, and backup diesel.
Projects like Carrington show that long-duration storage is no longer theoretical. The question now is which regions and organizations will move first, shape the market design, and lock in the long-term advantages of a genuinely flexible, low-carbon energy system.