Procurement · August 19, 2026 · 9 min read
How to Procure VPP Capacity: An RFP Guide for Utilities and Large Loads
Buying VPP capacity is unlike buying generation: you are contracting a probability distribution of delivery from consumer-owned devices. The procurement process has to price availability, not nameplate. This is the structure buyers use when the megawatts have to count for planning.
Step 1 — Define the product, not the technology
- Which hours must the resource deliver, and how many events per season?
- Notification lead time: day-ahead, hour-ahead, or 10-minute.
- Duration per event and maximum consecutive days.
- Locational requirement: system-wide, zonal, or a specific distribution feeder.
- Measurement basis: metered delivery versus baseline estimation.
Specifying "batteries" instead of "4-hour, day-ahead-notified, feeder-located capacity" is the most common scoping error and eliminates otherwise qualified bidders.
Step 2 — Normalize capacity claims before you compare
| Claimed basis | What it means | Discount to apply |
|---|---|---|
| Installed portfolio | All hardware ever installed | Heavy — most is not enrolled |
| Enrolled capacity | Devices signed up | Moderate — availability varies |
| Nominated capacity | Amount bid into a program | Light |
| Peak dispatchable | Delivered under peak conditions | Minimal |
| Measured event performance | Settled delivery from past events | None — this is the gold standard |
Require every bidder to restate their capacity in the same basis, and to attach the evidence. Directory profiles here record capacity basis explicitly for exactly this reason.
Step 3 — Evaluation criteria that predict delivery
- Measured delivery ratio across the last two event seasons in a comparable climate zone.
- Enrollment pipeline evidence: installer partnerships, channel, and cost of acquisition.
- Hardware diversity — single-OEM fleets carry concentrated firmware and vendor risk.
- Telemetry and settlement capability, including data formats your systems can ingest.
- Customer experience track record: opt-out rates, complaint volume, and churn.
- Balance sheet capacity to absorb non-performance penalties.
Step 4 — Structure the contract around performance
- Pay availability plus performance rather than enrollment counts.
- Set a firm delivery obligation with a measured ratio floor and a penalty schedule.
- Include a ramp schedule for the first season; new fleets rarely hit full obligation in year one.
- Require monthly telemetry deliverables and an annual independent measurement review.
- Define data ownership and what happens to enrolled customers if the contract ends.
Timeline and common failure modes
A realistic procurement runs three to six months from scoping to signature, plus a season for enrollment ramp. The failures repeat: contracting on nameplate, ignoring locational constraints, underestimating enrollment cost, and specifying telemetry the fleet physically cannot produce.
Frequently asked questions
How do you compare VPP aggregator bids fairly?
Restate every bid on the same capacity basis, weight measured event delivery from past seasons above enrolled nameplate, and price availability separately from energy delivered.
How long does VPP capacity procurement take?
Typically three to six months from scoping to contract signature, followed by one enrollment season before the fleet reaches its full contracted obligation.
What contract terms matter most for VPP capacity?
A measured delivery obligation with a penalty schedule, a first-season ramp allowance, defined telemetry and settlement deliverables, and clear treatment of enrolled customers at contract end.
Find the operators serving your market
Compare source-linked VPP aggregators and flexible capacity providers by state, technology, and program status.
Keep reading
Virtual Power Plant Companies: Who Runs VPPs in North America
Aggregators, OEM fleets, DERMS platforms and utility programs all call themselves VPPs. Here's the actual market structure and who sits in each layer.
How VPPs Are Being Used to Absorb Data Center Load Growth
Hyperscaler load is arriving faster than turbines and transmission. Flexible distributed capacity is the only resource that can be procured on that timeline.