Integrated Storage

Battery energy storage for data centres

Our 100+ MW BESS systems are co-located with on-site generation, ensuring seamless power delivery and backup. Paired with our CCGT and solar facilities, we deliver 24/7 reliability without grid dependency while generating revenue from grid services.

100+ MW
Storage Capacity
4 Hours
Discharge Duration
<100ms
Response Time
99.99%
Uptime Target

Why Battery Storage?

Critical for AI Infrastructure: AI workloads demand uninterrupted power. A brief outage can corrupt training runs worth millions of dollars. BESS provides seamless backup power, bridging any gap in primary generation and ensuring continuous operation.

Enabling Renewable Integration: Solar and wind are increasingly cost-competitive, but their variability creates challenges for continuous compute workloads. BESS smooths this variability, storing excess renewable generation and discharging when needed.

Revenue Generation: Beyond backup power, BESS generates revenue through grid services: frequency regulation, peak shaving, and capacity payments. In markets like Ontario and Alberta, these services can offset 20-40% of system costs.

Key Capabilities

100+ MW Storage Capacity

Grid-scale battery systems provide massive energy storage for peak shaving and backup power.

4-Hour Discharge Duration

Long-duration storage ensures reliable power delivery during extended peak demand periods.

Frequency Regulation

Millisecond response times for grid frequency regulation services, earning premium compensation.

Revenue from Grid Services

BESS systems generate revenue through ancillary services, capacity payments, and energy arbitrage.

Grid Services

Frequency Regulation

Battery systems respond in milliseconds to grid frequency deviations, providing critical balancing services that maintain grid stability.

Premium compensation for fast-response services

Peak Shaving

Store low-cost energy during off-peak hours and discharge during peak demand, reducing grid strain and earning arbitrage revenue.

Reduce peak demand charges by 30-50%

Renewable Integration

Smooth intermittent solar and wind output, storing excess generation and providing power when renewable output drops.

Enable higher renewable penetration

Backup Power

Provide uninterruptible power supply for critical AI workloads, ensuring 99.99% uptime even during grid outages.

Enterprise-grade reliability

Alignment with Canada's Net Zero Goals

Clean Electricity Regulations

BESS systems are essential for integrating variable renewables into the grid, directly supporting Canada's 2035 clean electricity goals.

Critical enabling technology

Provincial Renewable Targets

Alberta, Ontario, and other provinces are actively procuring grid-scale storage to support renewable energy expansion.

Active procurement programs

Grid Modernization

BESS supports Canada's grid modernization efforts, providing the flexibility needed for a decarbonized electricity system.

Infrastructure modernization

Federal Clean Tech Incentives

Battery storage projects qualify for accelerated capital cost allowance and clean technology investment tax credits.

Favorable tax treatment

Greco Energy's Approach

Led by Leo Paskalidis, BESS is integrated into every Greco Energy facility, providing both operational resilience and revenue diversification. Our systems are sized for 4-hour duration, providing enough backup for maintenance windows and grid events while maximizing revenue from grid services.

100+ MW CapacityGrid Services RevenueRenewable Integration99.99% Uptime

Which chemistry, and where each one wins

Most suppliers sell one chemistry and argue it fits everything. A developer designing the whole campus can put each where it is genuinely better, so the honest comparison is the one below — including the parts that do not favour us.

Lithium-ion (LFP)Vanadium redox flowSodium-ion
Best atHigh power, fast responseLong duration, daily cyclingCost and cold weather
Typical duration1–4 hours4–12 hours2–4 hours
Cycle degradationDegrades with cycling; needs augmentationEffectively none — electrolyte does not wear outGood, still being proven at scale
Energy densityHigh — compact footprintLow — needs roughly 3–5× the spaceLower than lithium
Thermal runaway riskReal, and the reason NFPA 855 existsNone — aqueous electrolyte will not burnLower than lithium
Maturity for data centresProven and bankable todayProven, fewer deploymentsEarly — pilots, not fleets
Where GRECO uses itRide-through and fast grid servicesDaily load-curve cycling and heat-linked dutyWatching, not yet deploying

Indicative characteristics for utility-scale systems. Real selection depends on the site's load curve, footprint, fire code and offtake — which is the engineering conversation, not a table.

Long-duration storage, and why a data centre is an unusual case

Most long-duration energy storage is justified against renewable intermittency — storing solar through the evening. A data centre is a different animal: its load is close to flat and runs continuously, so the value of duration comes from arbitrage against demand charges, riding through longer grid events, and holding firm output while on-site generation ramps.

That is why the vanadium flow half of the stack matters here more than it would on a solar farm. A flow battery cycled hard every single day does not lose capacity the way lithium does, and a data centre is the rare load that genuinely will cycle it every single day.

Systems are delivered as containerised BESS units, which is what makes the capacity modular — blocks are added as load arrives rather than committing to the full bank on day one. Combined with on-site generation and controls, the result operates as a campus microgrid that can island from the utility and keep running.

Are battery energy storage systems safe?

The honest answer is that lithium-ion carries a real thermal-runaway risk, and pretending otherwise is how projects lose a permitting hearing. Runaway is a self-sustaining reaction in which a failing cell heats its neighbours until they fail too — which is precisely why NFPA 855 exists, and why spacing, deflagration venting, gas detection and fire-service access are designed in rather than bolted on.

Two design choices reduce it materially. LFP chemistry is substantially more thermally stable than the NMC cells used in vehicles, and it is what utility-scale systems now use almost universally. And vanadium flow batteries have an aqueous electrolyte that cannot burn — for the portion of the stack carrying duration rather than power, the runaway question does not arise at all.

For a municipality assessing a campus, this is usually the question behind the question. We would rather answer it plainly and early.

Battery storage for data centres: common questions

Because storage does four jobs a UPS cannot. It firms behind-the-meter generation so on-site power rides through a trip, shaves peaks against demand charges, provides seconds-scale response to utility dispatch signals, and lets a campus flex its load rather than draw flat. In British Columbia that last capability is what turns an AI data centre from a grid burden into a grid asset.

Lithium-ion delivers high power in a small footprint, making it right for ride-through and fast grid services measured in seconds to minutes. Vanadium redox flow batteries decouple power from energy, so they add long-duration capacity — four to twelve hours — with effectively no cycle degradation, suiting daily charge-discharge against a data centre's load curve. Most vendors sell one; a developer designing the whole campus uses each where it wins.

Sizing follows the job rather than the building. Ride-through and power-quality duty typically needs 15–30 minutes at full campus load; peak shaving and demand-response duty needs two to four hours; firming intermittent on-site generation can need longer. GRECO models state of charge, dispatch response and thermal behaviour in the campus digital twin before any container is ordered.

Yes, through three stacked streams. Demand charge reduction cuts the utility bill directly. Grid services — frequency regulation, capacity and demand response — are paid by the utility or system operator. And arbitrage between on-peak and off-peak energy prices adds a third. Because these use the same asset at different times, a well-dispatched BESS earns from all three.

It helps materially. BC Hydro is allocating limited capacity competitively, and a load that can shed on command is far easier to accommodate than one that cannot. Storage plus dispatchable compute lets a campus commit to reducing draw during system peaks, which strengthens an allocation case and can reduce the firm capacity the project needs to request in the first place.

A BESS is a grid-connected assembly of battery cells, power conversion equipment and controls that stores electrical energy and returns it on demand. At utility scale it arrives as containerised blocks: battery racks, an inverter or power conversion system, thermal management, fire detection and suppression, and an energy management system that decides when to charge and discharge. The battery itself is often the least interesting part — the controls and the power electronics are what determine whether it earns its keep.

Differently from a UPS, which exists only to bridge the seconds before a generator starts. A grid-scale BESS in a data centre campus does four jobs at once: it firms behind-the-meter generation so output stays flat while a plant ramps, it rides through grid disturbances, it discharges during expensive peak windows to cut demand charges, and it responds to utility dispatch signals by shedding or absorbing load within seconds. The energy management system arbitrates between those jobs continuously, because they occasionally want opposite things.

Lithium-ion carries a genuine thermal-runaway risk, which is why NFPA 855 governs spacing, deflagration venting, gas detection and fire-service access. Two choices reduce it substantially: LFP chemistry is far more thermally stable than the NMC cells used in electric vehicles and is now near-universal in utility-scale systems, and vanadium flow batteries use an aqueous electrolyte that cannot burn at all. We would rather state the risk plainly than have a permitting hearing discover it for us.

Not yet, and we would be cautious of anyone saying otherwise. Sodium-ion is genuinely promising — cheaper materials, no lithium or cobalt supply exposure, and notably better cold-weather performance, which matters in Canada. But deployments are pilots rather than fleets, and the bankability and warranty terms a project financier needs are not there. We track it and expect to deploy it eventually; we do not specify it into projects today.

Any single number quoted without a load curve is marketing. Cost is driven by power rating in MW, duration in hours, chemistry, site works, interconnection and fire-code compliance — and the last two surprise people most often. The more useful framing is what the system earns: demand-charge reduction, avoided firm capacity in an allocation request, grid-service revenue, and the generation capacity you do not have to build because storage covers the peak. We size against that, not against a price per kilowatt-hour.

Sources & Further Reading

International Energy Agency• 2024

Grid-Scale Energy Storage: Technology and Market Outlook

Canada Energy Regulator• 2024

Canada Energy Regulator: Energy Storage

U.S. Energy Information Administration• 2024

Battery Storage in the United States: An Update