Everything people ask before, during and after a run. If your question is not here, email vsa@powercurve.dk.
AeroVista estimates how much annual energy your turbines lose to blade damage, using inspection data you already have.
For every turbine in your workbook: the percentage of annual energy production lost, for the rotor and for each of the three blades. You also get a picture of each blade showing where the damage sits and how severe it is.
Anyone who owns, operates or inspects wind turbines and has blade condition data. You do not need to be an aerodynamicist to use it.
Nothing. Each organisation can have 25 turbines analysed in total, shared by everyone at its email domain. After that, email vsa@powercurve.dk and we will talk about what you need.
Power Curve ApS, an independent wind energy consultancy in Aalborg, Denmark.
You need two things: your details, and an email address you can receive mail at.
Yes: a recording of one whole analysis, from entering your details to downloading the results. It is under How it works on the AeroVista page, as download a video of a full analysis.
So we know who is using AeroVista and can help when someone gets in touch. The details are kept with the record of your run. What we hold, and for how long, is in Your data below and in the privacy notice on the page.
Because your results are sent to that address, and because the free allowance follows it. Confirming that you can receive mail there keeps both honest. It is not an account, and there is no password.
Once every 30 days in the same browser. If you clear your browser data or move to another computer, you are asked again.
It comes from noreply@aerovista.io, usually within a minute. Check your junk folder. If your organisation filters unknown senders, ask IT to allow that address. A code is valid for 10 minutes, and you can ask for up to three per hour.
That check is Cloudflare's, and it decides whether a visitor is a person rather than a script. Reload the page and try again. A corporate VPN or a strict network is the usual cause; if you can, try without the VPN.
Start from the input template, which you can download on the page. It has the right sheets and columns, dropdowns for the turbine model and rated power, and it fills in each turbine's measurement grid once the model is chosen. That is most of what used to go wrong.
Two sheets per wind farm site. A site sheet holding the damage data, for example Site1, and a list sheet named after it, Site1_TurbineList. You can have as many sites in one workbook as you like, as long as every site sheet has its list sheet.
One row per turbine, under three headers: TurbineNumber, TurbineMakeAndModel and RatedPower. Use the dropdowns rather than typing.
To add a turbine, type it on the next empty row and choose its model, and its columns appear on the site sheet by themselves. There is room for 200 turbines per site. To remove one, clear its three cells, and its severities on the site sheet.
Four columns per turbine: the distance from the hub, then the damage severity for blades A, B and C. The template has them ready for every row of the turbine list: they label themselves when you name the turbine, and the first fills itself in when you choose its model. That column has been called Radial Position and Blade Position before now, and a workbook using either name is read exactly the same way.
Each turbine's severities sit in the columns that belong to its row of the list. Sorting the list, or inserting or deleting a row, could leave one turbine's damage under another turbine's name, with nothing on screen to show it. So both sheets are protected against those, and against typing over the distances. You can still type, paste, choose from the dropdowns and widen columns. To remove a turbine, clear its cells instead.
The measurement grid for that turbine, and the template fills it in for you once you have chosen the model. It steps 0.25 m at a time from the centre of the hub out to the blade tip: 2 to 68 m for a V136, 1.5 to 46.5 m for an SWT-2.3-93. The last step is sometimes shorter than the rest, so the grid ends exactly on the tip rather than just short of it: a GE1.5-77 with a T37 blade finishes 38.10, 38.35, 38.50. Change the model and the grid rebuilds itself, so choose the model before entering severities: a change afterwards moves the distances beside them.
That is the way inspection reports give damage locations, so a photo marked "Location: 64 m" goes on the row that says 64, with no arithmetic. Put each blade's severity beside it, and 0 on every row with no damage.
No. The template's grid is already measured from the centre of the hub, the same end your report measures from, so the numbers go straight in. AeroVista converts to distance along the blade at its own end, and the run report inside your results says how many turbines it did that for.
A workbook measured along the blade instead, from 0 at the root, is read correctly too. AeroVista tells the two apart from the numbers themselves, not from the column heading: a point cannot be 68 m along a 66 m blade.
How badly that piece of the blade is eroded, as a whole number from 0 to 5. Every position on every blade needs one, so a blade with nothing wrong with it is a column of zeros.
| Severity | What the blade section looks like |
|---|---|
| 0 | No visible damage |
| 1 | Small individual pits, or light dirt and bug impact |
| 2 | Distributed shallow pinholes, medium dirt and bug impact, or the leading edge protection starting to peel |
| 3 | Shallow pits merging into small patches, heavy dirt and bug impact, or the leading edge protection peeled off |
| 4 | Damage reaching back from the leading edge, a mixture of shallow and deep |
| 5 | No paint or gel left, bare laminate showing, with a defined step |
Scoring from words alone is hard, so there is a guide with example photographs of each severity level.
From your blade inspection, usually a report or images from an inspection provider. Turning what an inspector saw into one of those numbers is a judgement, so score against the photographs in the damage severity guide rather than from memory. Scoring the same way every time matters more than scoring perfectly: it is what makes two inspections of the same turbine comparable.
AeroVista supports 25 models, listed on the page under Show accepted turbine models. Adding one needs blade and airfoil data for it, so email vsa@powercurve.dk and tell us which model you need.
Yes, but both spellings work. The template and the page now show names like GE2.8-127 (LM62 blade); older workbooks written as GE2.8-127_LM62 are still read correctly. A name that does not say which blade, like GE2.8-127 on its own, is refused rather than guessed at.
.xlsx only, up to 50 MB. If yours is an older .xls, open it in Excel and use Save As to make an .xlsx. CSV files are not accepted.
You can put a maximum of 200 turbines in one workbook, spread across one site or several.
The wind at your site decides how much the damage costs you, so AeroVista asks before it calculates.
The wind climate dictates how long a turbine spends at that mean wind speed. The lower the mean wind speed, the more time the turbine spends in the part load region where effects of erosion are higher, and the greater the AEP loss.
The preset closest to your site's annual mean wind speed at hub height: 10, 8.5 or 7.5 m/s. All three assume a Weibull shape factor of 2, which is the usual assumption. 8.5 m/s is the default.
Choose the last option and enter the annual mean wind speed, between 4 and 11 m/s, and the Weibull shape factor, between 1 and 5.
It describes how spread out the wind speeds are over a year, rather than how strong they are. Most sites sit between 1.8 and 2.5. Use 2 if you do not have a measured value.
The wind conditions used are printed in the run report inside your results package, so a result always says what it was calculated against.
Upload the workbook, press Run AeroVista, and the page follows the run until it finishes.
You can run up to 25 turbines and get the results in minutes.
Yes. The run carries on without the page, and the results are emailed to the address you verified. The same is true if your connection drops: the page says so, and the run keeps going.
Turbines in workbook is what you uploaded. Turbines calculated climbs as they finish. Turbines skipped is everything AeroVista could not use, and the Warnings box says why, turbine by turbine.
It is optional, and it is the company whose turbines this analysis is for, such as the owner or operator. It is printed in the run report inside your results, which helps when you are analysing fleets for more than one owner.
Press Clear form. It clears the workbook, the wind farm owner and the wind conditions, and keeps your details, so you can go straight into the next one. To change your details, press Edit user information.
Under Recent runs at the bottom of the page. That list is stored in your browser only, so it is gone if you clear your browser data or move to another computer. Your results email is the copy that lasts.
A finished run gives you a ZIP file, on the page and by email.
| File | What it holds |
|---|---|
AEP_Loss_Summary.xlsx |
One sheet per site, one row per turbine: the Turbine ID, then the AEP loss for each of the three blades, then the Rotor AEP Loss, the same loss in MWh a year, and the turbine's clean AEP |
Images/HeatMaps/<site>/ |
One picture per turbine, showing the damage along its three blades |
Images/ |
Overview plots for the site |
AeroVista_Run_Report.txt |
What was run: the file, the wind conditions, the turbine counts, the warnings, and how many turbines were measured from the hub |
The amount of energy that a turbine loses because of the damage you described, as a percentage. It is the sum of the three blade losses. A value of −2.6 % means the turbine is expected to produce 2.6 % less energy over a year than the same turbine with clean blades.
Rotor AEP Loss (MWh) is the same loss in energy: the turbine's clean AEP, in the wind conditions you chose, times its Rotor AEP Loss. The page prices the whole run's loss at a power price you can change between 10 and 100 EUR/MWh, starting at 50. The energy is before availability, wake and electrical losses, so it is somewhat more than a turbine actually sells.
The part of that loss belonging to each blade.
Each picture shows one turbine's three blades, with the loss contribution to rotor AEP loss coloured along them. Areas marked in red contribute the most to the overall AEP loss and areas marked in green contribute the least to the overall AEP loss.
Not directly. It is lost production. After a run, the page turns it into revenue lost each year, at a power price you set on its slider.
The download link works for 24 hours, and the files are deleted after 30 days. It is named for the date and time you ran it, in your own time zone, so several of them sort into the order you ran them. Save it somewhere of your own when you get it, and the results will also be emailed to the registered email address once the run is completed.
A turbine AeroVista cannot use is skipped, and the rest of the run carries on. The Warnings box names the site and the turbine and says which of these it was.
| What the warning says | What to change |
|---|---|
| Unsupported model | Use the dropdown in the template. |
| Rated power the model does not have | Pick one of the ratings the warning lists for that model |
| Severity values outside 0 to 5, or not whole numbers | Use whole numbers from 0 to 5. |
| Non-numeric severity or radial values | Take text out of the data columns. 3 - moderate is text; 3 is a number |
| Radial spacing that is not 0.25 m | The template builds this column and locks it, so this usually means something was typed over it with the protection switched off. Clear it and choose the turbine model again to rebuild it. Building it by hand, step 0.25 m the whole way, except the last step onto the tip, which may be shorter. A missing row in the middle is the usual cause: fill it in with severity 0 rather than leaving it out |
| Radial positions out of order, or repeated | Sort the whole block by radial position, severities included, and delete the repeat |
| Maximum radial position past the blade length | Check the model is right. Positions measured from the hub are converted for you when the whole blade is covered, so what is usually left is the wrong model, or a block that starts partway along the blade |
| A turbine in the list with no data block, or a block with no list row | In the template, add turbines on the next empty row, never between two others. In a workbook built by hand, add the missing one or remove the other |
| A turbine listed twice | Keep each turbine on one row, and clear the other |
| Damage severities where the blade has already ended | Usually the model was changed after the severities went in. Check the model, then move or clear the severities the template shows in red |
| Severities in columns with no turbine name | A turbine was cleared from the list after its severities went in. Clear those severities, or put the turbine back. No turbine is skipped for this one |
| A row with a model but no turbine name | Name the turbine, or clear the row |
A site with no _TurbineList sheet |
Add the list sheet. This skips the whole site, not one turbine |
No. Only turbines that produced a result count towards your 25 turbine allowance.
The status says No valid results generated, and the run cost you nothing. It is almost always one mistake repeated down the workbook rather than many different ones, so fix the first warning and the rest usually go with it.
Each organisation can have 25 turbines analysed, in total. It does not reset each month.
Yes. Everyone with an email address at your company's domain draws on the same 25, so if a colleague has analysed 10, there are 15 left for all of you. An address at a public email provider such as Gmail or Outlook has its own 25, because its domain says nothing about who it belongs to.
Only turbines that produced a result. A turbine that was skipped, a workbook that could not be read, and a run that failed on our side all cost nothing. Analysing the same turbine twice counts twice, because it is calculated twice.
The next upload is refused with a message asking you to email vsa@powercurve.dk. Nothing you already have is affected: your earlier results stay where they are.
The run is refused before anything is calculated, and the message says how many turbines the workbook needed and how many you have left. Nothing is counted, so you can split the workbook to fit, or email us.
Yes. Email vsa@powercurve.dk, tell us roughly how many turbines you are looking at and what you are using AeroVista for, and we will sort something out.
The page tells you. When you verify your email address it says how many of your 25 turbines are left, and whether they are shared with your colleagues, on the card with your details, and every finished run updates that number.
Because a run can be one turbine or two hundred, so counting runs would charge two people very differently for the same amount of work. Until 18 September 2026 the limit was 20 runs per 30 days, which did exactly that. From then until 27 October 2026 it was 50 turbines per email address.
Power Curve ApS is responsible for the personal data AeroVista processes. The full privacy notice is on the page; this is the short version.
| What | How long it is kept |
|---|---|
| Your workbook, results and heatmaps | 30 days, then deleted automatically |
| Your name, email, organisation, role, country and any notes | 36 months after your last run |
| A record that you accepted the notice and terms, and which version | With that run's record |
| Your IP address, as a keyed hash, for abuse protection | 30 days. The address itself is never stored |
Two service providers, both under data processing agreements. Microsoft Azure hosts AeroVista, stores the data in Europe and sends the email. Cloudflare runs the human check and may process your IP address and browser information. Nobody else, and we do not sell anything.
We may keep the blade condition data you upload and the results from it, and use them in aggregated or anonymised form to improve AeroVista and validate its models. We will not publish or share anything that identifies you, your organisation or a specific wind farm without your agreement. That permission is in the terms of use, because turbine data is commercial data rather than personal data.
No. Your details are not used for marketing.
Email info@powercurve.dk. You can ask for access, correction, deletion, restriction, objection or a portable copy, and we answer within a month. That address is for data questions; everything else goes to vsa@powercurve.dk.
Please do not, unless the analysis needs it. Turbine numbers and damage data are all AeroVista reads.
If a message shows a code such as AV-602, that code tells us exactly where it happened. Send it to us with the Run ID and we can look the run up.
Press Copy details for support, next to the Run ID in the result, and paste that into an email to vsa@powercurve.dk. It carries the Run ID, the status, the message with its code, your file name, the turbine counts and the warnings. Nothing else is needed.
| What you see | What it means |
|---|---|
| Workbook needs correcting | AeroVista could not read the workbook as an AeroVista input workbook. Start from the template and move your data into it. Nothing was counted |
| No valid results generated | The workbook was read, but every turbine was skipped. The warnings say why, turbine by turbine |
| Completed with warnings | You have results, and some turbines were skipped. Read the warnings before using the numbers |
| This run has not reported any progress for more than 10 minutes | The run stopped without saying so. Run it again, and tell us if it happens twice |
| Could not reach AeroVista | Your connection, or the site being down. Nothing was sent, so nothing was counted |
| AeroVista is unavailable right now | Email us and quote AV-805 |
Check the junk folder first: the results come from noreply@aerovista.io. The same package is on the page as long as the link is still valid. If a whole organisation sees nothing, ask IT to allow that address.
Only the page lost contact. The run carries on, and if it finishes the results are emailed to you.
Verify again. If it happens straight after verifying, tell us and quote the code shown with the message: AV-213, AV-222, AV-223 and AV-224 look similar to you but mean different things to us.
AeroVista gives engineering estimates from a model, not measurements from your turbines.
It takes the blade shape and the aerodynamic data for the model you named, applies the damage you described to the respective part of each blade, and works out how much power the rotor makes across the range of wind speeds your site sees in a year. The difference from clean blades is the AEP loss.
Your severity numbers, above everything else. The same blade scored two categories higher gives a noticeably larger loss. After that, the model and rated power being right for the turbine, and the wind climate being close to your site's.
The same workbook at the same wind conditions gives the same answer every time, so runs are comparable with each other. That makes it useful for ranking turbines and for tracking a fleet over time, which is what most people want it for.
Talk to us first. The results are estimates and depend on inspection data that we cannot check, so they are not a guarantee of any turbine's actual performance, and the terms of use say so.
Something changed: the severity data, the model or rated power, or the wind conditions. The run report in each package records what that run used, so comparing two reports usually finds it.
Two addresses, so your message lands with the right person.
| What you are writing about | Where to send it |
|---|---|
| An error, a code, a turbine that was skipped, a model we do not support, more turbines than your allowance, or anything about using AeroVista | vsa@powercurve.dk |
| Your personal data: a copy of it, a correction, deletion, or anything in the privacy notice | info@powercurve.dk |
When something went wrong with a run, include the Run ID and the error code, or simply paste what Copy details for support gives you. With those, we can find the run itself rather than asking you to describe it.
Power Curve ApS, Kastetvej 2, st., 9000 Aalborg, Denmark.