PowerVeritas
Services / S/01, FlagshipPowerVeritas Ltd

Wind Farm Performance Reviews

PowerVeritas performance reviews reconstruct wind farm production from turbine level SCADA and quantify every loss by cause.

Know how much your site is losing, why and which fixes pay back first, with optional ongoing monitoring against that baseline.

S/01.1

Initial deep diveFlagship

Performance Reviews reconstruct the full production history of the site by turbine and time period. Losses are attributed across wind, availability, curtailment and underperformance, and flagged turbines are investigated for control, sensor or component issues. A prioritised recovery plan identifies the interventions with the largest energy impact.

Scope · Initial deep dive
  • 01Production history & loss attributionWhy the site missed budget, broken down by cause.Site production is reconstructed from SCADA and losses quantified across wind, availability, curtailment and underperformance.
  • 02Data quality & reconciliationWhether the underlying data can be trusted.Completeness, consistency, reliability and cross-check of available SCADA data.
  • 03Power curve performanceWhich turbines are underperforming, by how much and where in the wind range.Bin level analysis corrected for density and turbulence, against site specific and OEM reference curves.
  • 04Curtailment quantificationHow much production is being lost to curtailment and to which cause.Grid, noise, shadow flicker and wake sector curtailment separated and quantified in MWh and revenue.
  • 05Engineering reviewWhat the data is telling you that headline KPIs miss.A Chartered Engineer reviews operational plots, cross-checks the data and identifies control faults, sensor issues and root causes hidden behind dashboard summaries.
  • 06Recovery planWhat to fix first and what it is worth.Prioritised list of interventions, ranked by production impact and implementation effort.
S/01.2

What you receive

Deliverable A

Executive waterfall report

Top line losses attributed, ranked and quantified. Written for boards, owners and lenders.

Deliverable B

Turbine level diagnostics pack

Per turbine findings with supporting charts and data, for asset managers and operators to act on.

Deliverable C

Data export

Underlying turbine level data and loss attribution outputs, delivered in Excel for integration with existing client reporting.

Optional
Monthly monitoring

Ongoing monthly monitoring add-on

A recurring service that tracks performance against the baseline, flags new issues and updates the recovery plan monthly.

S/01.3

From data to diagnosis

Most performance numbers fall apart under scrutiny because the underlying data has not been properly classified. The two illustrative outputs below are from a real review. SCADA is first classified into operating states so each 10 minute record is correctly categorised, allowing accurate loss quantification across curtailment, downtime, sensor faults and underperformance. A measured site curve is then produced from the clean data and compared against the contractual OEM curve.

Step 01 / Operating state classification
Per turbine state by 10 min interval
Heatmap of operating state by turbine and time. Most cells are blue (normal operation). Visible patterns include occasional grey bands (offline), turquoise vertical band on Turbine 11 (internal curtailment) and green horizontal stripes (external curtailment events affecting all turbines simultaneously).
Every 10 minute SCADA record is tagged as normal, curtailed (internal/external), startup/shutdown, offline or bad data. Only normal points enter the power curve fit. Patterns surface immediately. Here, Turbine 11 shows persistent internal curtailment and the site experiences periodic full fleet external curtailment.
Step 02 / Measured vs OEM power curve
Single turbine fit, classified scatter
Scatter plot of wind speed against power output for a single turbine. Most points form a tight band around the OEM curve. Off curve points are colour coded by cause. External curtailment (low power points), internal curtailment (capped points below rated), startup or shutdown and flagged bad data. The measured binned curve runs slightly above the OEM curve at low wind speeds and converges at rated.
With states classified, a binned measured curve is fit to normal operation points only. Off curve scatter is preserved and colour coded so curtailment, sensor faults and startup/shutdown losses are visible (and quantified separately in the loss waterfall) rather than silently distorting the fit.
S/01.4

What data is needed

A performance review runs on 10 minute SCADA at turbine level plus a small set of site facts. The tables below are everything that is required to start.

Site information, once at onboarding

Site name and location
Site export limit, MW
Turbine coordinates
Turbine model, rated power and rotor diameter
Turbine power curve

Turbine SCADA, 10 minute resolution

Timestamp, with time zone confirmed
Turbine ID
Active power, kW
Wind speed, m/s
Rotor speed, RPM
Pitch angle, degrees
Nacelle position, degrees

That set is enough to run a review. A short list of around ten further tags is shared with the proposal. They are optional and they sharpen the attribution rather than enable it. If you can supply only one, make it the power setpoint, which materially improves how cleanly curtailment is separated from underperformance.

CSV or Excel, uploaded to a secure SharePoint folder or by the method set out in your Scope of Work.

Common questions

What does a wind farm performance review cover?

A performance review reconstructs the full production history of the site by turbine and time period from SCADA data. Energy losses are then quantified by cause across wind resource deviation, availability, curtailment and turbine underperformance. Power curves are analysed for control faults, sensor drift and component issues. The output is a prioritised recovery plan ranking interventions by energy impact and implementation effort. Reviews are available as a one off or with ongoing monthly monitoring against the established baseline.

What SCADA data is needed for a performance review?

Seven tags at 10 minute resolution, turbine level. Timestamp with the time zone confirmed, turbine ID, active power, wind speed, rotor speed, pitch angle and nacelle position. Alongside those, a short set of site facts, the site name and location, export limit, turbine coordinates, turbine model with rated power and rotor diameter, plus the turbine power curve. A further list of around ten optional tags is shared with the proposal, which sharpen the attribution rather than enable it.

What if some of my SCADA signals are missing?

Most reviews proceed. The seven required tags are the core. Beyond those, missing signals narrow what can be concluded rather than stopping the work. The report says explicitly what could not be assessed and why. Data completeness is checked and reported as a workstream in its own right.

How is this different from a monitoring platform?

A platform tells you what happened and leaves the engineering review to you. A performance review does that review for you. Every loss is quantified in MWh, attributed to a cause and ranked into a recovery plan by production impact and implementation effort. You get the conclusions and the actions rather than a dashboard of alerts and a job still to do.

Will the findings stand up if my O&M contractor disputes them?

They are built to be examined. Every loss is traceable from raw input data through to the headline number, and diagnostics sit alongside conclusions so the working can be checked. Losses only enter headline figures where the data supports them. The conservative position is deliberate, because a number that survives challenge is worth more than a larger one that does not.

Do I have to commit to monthly monitoring?

No. Monthly monitoring is optional and most clients decide after the initial review whether tracking the baseline is worth it. It runs under its own Scope of Work with a notice period, set out in the agreement before it starts.

What drives the cost of a performance review?

A performance review is priced on a formula, a fixed site fee plus a per turbine fee, so the cost follows turbine count rather than hours worked. For a ten turbine site that is a low four figure engagement. Monthly monitoring afterwards is optional and priced separately, at around a fifth of the initial review fee each month. Sites above about 40 turbines are quoted individually. Pricing is indicative, with the fee fixed in the proposal before work starts.

What do I actually get at the end?

Three things. An executive waterfall report with losses attributed, ranked and quantified, written for boards and lenders. A turbine level diagnostics pack with the supporting charts and data for your asset managers to act on. And a data export in Excel so the loss attribution can feed your existing reporting.

Want to review a site or portfolio?

Book a call → Email PowerVeritas