Data Center BESS | Interconnection, Permitting & Revenue

AI FACTORIES & DATA CENTERS

Turn Your Power Constraint Into a Grid Asset

Grid operators are tightening interconnection standards for large loads. We design and permit the BESS strategy that gets your campus connected—faster, compliant, and positioned to earn from the grid.

The Power Playbook Has Changed

The interconnection landscape for large loads has fundamentally shifted. Grid operators are capacity-constrained. Regulators are demanding that new data centers prove they won’t degrade system reliability. The traditional approach (file an application, then wait for utility upgrades) now means waiting five to seven years in many markets.

Energy storage offers a way forward. But a BESS co-located with critical computing infrastructure is not an oversized C&I project. The duration sizing, fire code requirements, augmentation strategy, and controls architecture are fundamentally different—and getting them wrong means rejected permits, delayed campuses, and stranded capital.

NVIDIA Now Requires Qualified BESS for AI Factories

In May 2026, NVIDIA published formal BESS Self-Qualification Guidelines for AI data center campuses. The framework defines 12 performance tests covering AI load buffering, grid-adaptive operation, ride-through, demand response, and black start, plus business and supply chain readiness requirements including demonstrated manufacturing throughput and a credible plan to scale production tenfold within 24 months.

The framework validates the equipment. It does not validate that the BESS can be permitted, sited, interconnected, or commissioned at a specific site. NVIDIA explicitly excluded fire protection, deferring entirely to the local Authority Having Jurisdiction and NFPA 855. A BESS that passes all 12 NVIDIA tests can still fail to get a building permit.

Carina has integrated NVIDIA’s framework into our OEM evaluation process. When we run the Technical RFI for a data center engagement, the RFI includes questions mapped to NVIDIA’s core requirements. The Design Basis Memo documents the selected OEM’s qualification status. And the development disciplines NVIDIA intentionally excluded from their scope, fire protection, interconnection, and local permitting, are the disciplines Carina manages.

NVIDIA qualifies the equipment. We develop and permit the project.

What a Strategic BESS Deployment Enables

Skip the Interconnection Queue

Co-located storage that provides dispatchable capacity transforms your facility from a grid problem into a grid solution.

In Oregon, this approach brought a 100+ MW campus online years ahead of schedule, and in PJM, proactive storage is becoming an interconnection prerequisite.

Reduce Your Genset Capital Stack

Data center gensets are routinely oversized 30 to 50 percent above steady-state load to handle motor starting currents from chillers, compressors, and pumps.

A BESS on the critical power bus absorbs the inrush transient, allowing gensets to be sized closer to steady-state load reducing your capital stack

Unlock Front-of-Meter Revenue

Sized beyond facility backup, your BESS becomes a revenue-generating asset in energy, frequency regulation, and capacity markets.

The sophisticated approach is a hybrid BTM/FTM configuration—but that requires navigating dual regulatory frameworks most developers aren’t set up to manage alone

Future-Proof Controls and Compliance

NVIDIA’s BESS Self-Qualification Guidelines now define specific telemetry, controls transparency, and grid-adaptive requirements for AI factory storage: real-time reporting at one-second polling, EMT-validated control models, and documented ride-through behavior.

We evaluate OEM compliance during the Technical RFI, not after commissioning.

The Carina Model for Data Center BESS:

Carina Energy is not an equipment vendor or EPC contractor. We are your Owner’s Representative—managing the technical team on your behalf so that professional liability stays with insured specialists, not with you.

1. Hyper-Local Execution + Expert Overlay

We hire local civil and environmental firms who have relationships with your AHJ. Then we layer national BESS and fire protection specialists on top. You get local speed with gigawatt-scale rigor

2. OEM-Agnostic, Fire-Code-First

We use a two-stage RFI/RFP process: technical selection first (validated by our FPE against local fire code), commercial competition second. We never lock lithium pricing during design.

3. Liability Separation + Insured Accountability

Carina never stamps drawings. Engineering liability sits with the insured Engineer of Record and FPE—not with you, and not with us. We manage the process; the specialists carry the professional responsibility.

Seven ways we help you create value from co-located BESS, beyond backup power

Every engagement starts with a Use Case Mapping exercise. We identify which of these seven use cases apply to your site, which stack, and which conflict.

Interconnection Acceleration

Size BESS to avoid a material modification trigger and bring load online ahead of firm interconnection capacity.

Capacity Factor Improvement

Extract more usable capacity from a fixed POI by firming variable load profiles and managing peak demand.

Front-of-Meter Revenue Stacking

Participate in ISO ancillary services, energy arbitrage, and capacity markets during periods when BTM functions are not called.

Time-of-Use Arbitrage

Charge during low-price periods, discharge during high-price periods, reduce net electricity cost per GPU hour.

Demand Charge Management

Extract more usable capacity from a fixed POI by firming variable load profiles and managing peak demand.

Genset Capacity Reduction

Absorb inrush transients to allow gensets to be sized closer to steady-state load. CapEx reduction on your backup generation stack.

Shared O&M and Monitoring

Your existing 24/7 NOC monitors BESS through SCADA integration, eliminating a separate monitoring contract.

Data Center Segments We Work With

Crypto-to-AI Site Conversions

Companies converting mining operations to AI compute hosting. You have the land and the interconnection.

We bring the BESS development capability to make the conversion work.

Regional Colocation Developers

The 10 to 100 MW tier. Small to medium size development teams squeezed in interconnection queues by hyperscaler demand.

Co-located BESS accelerates your timeline and differentiates your site.

Powered Land Developers

Building 20 to 100 MW powered sites for sale to colo operators and AI hosting companies.

BESS developed against NVIDIA’s framework makes your site more valuable to the buyer

AI-Native Compute Providers

GPU-as-a-Service (neoclouds) companies racing to deploy capacity. Built around chip access, not power infrastructure.

We provide the BESS development and permitting team you have not hired, and don’t need.

Supply Chain Compliance Runs Deeper Than You Think

NVIDIA requires OEMs to disclose supply chain risks, including export control exposure and geographic concentration. Federal tax credit rules under the IRA impose a separate screening obligation on the same OEMs: battery cells from designated foreign entities of concern can disqualify the entire project from the Section 48E Investment Tax Credit.

We evaluate both NVIDIA’s business readiness requirements and IRA foreign entity of concern provisions during OEM selection, so your vendor decision survives both the technical qualification and the tax credit analysis.

Learn more: FEOC Compliance for BESS