Quick answer
A contractual commitment to meet a defined plant or equipment availability metric, with specified period, capacity and time boundaries, event treatment and remedies.
What Is an Availability Guarantee?
Answer: An availability guarantee is a contractual commitment to meet a defined level of equipment or plant availability over an agreed period. Its meaning depends on the calculation, capacity boundary, eligible hours, event categories and exclusions. It does not automatically guarantee annual energy or revenue, and any payment, cap or remedy follows the actual agreement rather than a universal industry rule.
A percentage without its definition is not enough to compare offers. Two contracts using “99% availability” can count different hours, equipment and events. O&M service scope, an equipment warranty and a production guarantee are related but distinct commitments.
Define the Measurement First
The NREL/Sandia/SunSpec 2018 O&M best-practices report, Appendix F, presents model availability terms. It separates availability from energy performance, addresses partial capacity loss and treats event categories and exclusions as matters to define. The example is guidance, not your executed contract or a universal guaranteed threshold.
For a project review, request the following specification before comparing the headline percentage:
| Question | Required clarification |
|---|---|
| What asset is covered? | Whole plant, inverter, block or another defined scope |
| Which capacity reference? | AC/DC rating and weighting of partially unavailable equipment |
| Which hours are eligible? | Calendar hours or defined production opportunity, with the detection criterion |
| Which period? | Month, contract year or another period; commencement and partial periods |
| Which events count? | Explicit classification and approval process for each exclusion |
| What is the evidence? | Data sources, timestamps, event records and missing-data treatment |
| What follows a shortfall? | Calculation, claim procedure, remedy, caps and dispute process |
These questions prevent an apparent precision that the underlying data or agreement does not support. Professional review should resolve wording and enforceability in the relevant jurisdiction.
Time-Based Arithmetic
For a hypothetical binary whole-plant metric, assume every hour in a non-leap calendar year is eligible, no exclusions, and 200 hours of complete unavailability:
Availability = (eligible hours − unavailable hours) ÷ eligible hours × 100.
(8,760 − 200) ÷ 8,760 × 100 = 97.7169%.
Against an assumed 98% target, the unrounded shortfall is 0.2831 percentage points. That calculation alone does not establish a payment: the agreement must specify treatment of rounding, shortfalls and remedies.
Under those same narrow calendar-hour assumptions, 99% permits 87.6 hours of counted downtime. If the denominator instead contains 4,000 eligible production hours, 99% permits 40 hours. Neither allowance is universal. A leap year also changes a calendar-hour denominator.
A loss of telemetry is not proof of a complete plant outage. Establish whether the equipment was unavailable using the agreed evidence and missing-data procedure.
Partial Outages Need a Capacity Boundary
Consider an independent hypothetical 1,000 kW plant, 100 eligible hours, and one event that makes 200 kW unavailable for 10 hours. With no overlapping outages or exclusions:
Equivalent full-plant unavailable hours = 10 × 200 ÷ 1,000 = 2 hours.
Capacity-weighted availability = (1 − 2 ÷ 100) × 100 = 98%.
Counting the entire plant as down for all ten hours would instead give 90%, misrepresenting this declared partial-capacity example. This calculation excludes gradual degradation and does not establish how much energy was lost.
Availability Is Different from Energy and Revenue
Time and capacity weighting do not account for all irradiance or price differences. PVsyst’s unavailability documentation explains that simulated outage energy loss depends on the timing, season and weather, so its energy-loss percentage need not match the time fraction.
An energy-weighted metric needs an agreed expected-energy method, meter boundary and rules for allocating loss to unavailable equipment. A tariff-weighted financial measure adds another boundary. Do not call energy weighting “revenue weighting” without that extra calculation.
A system can be available yet underperform because of another loss. Conversely, downtime in a low-resource interval can have a different energy effect from equally long downtime at peak production. Avoid counting the same outage loss again under separate performance guarantees.
Event Records and Commercial Handoff
Retain each event’s asset, start/end timestamps, affected capacity, cause, evidence, repair action and proposed classification. Record overlapping outages without double counting. Distinguish communications failures, grid restrictions, maintenance and equipment faults; inclusion or exclusion needs the actual terms, not a blanket list.
Designers should consider safe access, isolation and maintainability with the responsible engineers. Service teams should align monitoring and spare-parts arrangements with the promised scope. Repair-speed metrics such as MTTR help operational review but do not replace the agreed availability calculation.
Sales teams should provide the definition and remedy alongside the percentage. Do not translate a time-based target directly into guaranteed dollars, assume an automatic damages rate, or describe a contractual target as a measured operating history. Solar warranties require their own scope and conditions.
When evaluating SurgePV, ask which outage, energy and financial assumptions the proposed workflow supports. A software estimate does not itself issue an O&M guarantee.
Common Questions
What is a good availability guarantee?
Evaluate the scope, eligible hours, evidence, exclusions and remedies with the headline target. No percentage is automatically suitable for every plant or contract.
Is it the same as a production guarantee?
No. Availability concerns defined operational capability; a production guarantee concerns an agreed energy-performance commitment. Read both definitions and avoid duplicate loss claims.
How is availability verified?
Use the agreed equipment/event data and calculation, including partial outages and missing-data rules. Zero output or missing telemetry alone does not identify a covered equipment failure.
Related Glossary Terms
About the Contributors
CEO & Co-Founder · SurgePV
Keyur Rakholiya is identified by SurgePV as its CEO and a company co-founder. His SurgePV author page lists only role information that can be tied to the public profile below; credentials, project totals, testing claims, media appearances, and speaking engagements are not asserted without retained evidence.
Editorial contributor · SurgePV
Rainer Neumann is credited as an editorial contributor on SurgePV content. This profile does not assert engineering credentials, project totals, software-testing experience, education, speaking engagements, or media citations because independent verification evidence is not retained in the publication record.
