The dramatic growth in turbine hub heights over the past two decades is one of the most visible signs of the wind industry’s evolution. To understand the operating environment at these new heights – now commonly exceeding 100 meters – developers and owners have looked to innovative wind measurement technologies, such as remote sensing.
Ground-based SoDAR and LiDAR devices provide rich wind measurement datasets across the entire turbine blade sweep. Because they are also mobile and easily deployed, they have become a widely used alternative to met towers to help reduce the costs, logistics and permitting challenges related to using tall towers. The past five years have also seen growth in the use of remote sensors at operating wind parks for anemometry validation, wake studies and diagnosing performance issues.
This last area is a key challenge for the industry because the process is extremely costly yet essential for operators looking to remedy underperformance concerns and derive as much value as possible out of existing assets. Thus, the new standard for power performance testing (IEC 61400-12- 1, released in April) has been much anticipated. It supersedes the previous 2005 version and includes a number of updates. Those related to remote sensing have drawn the most attention – and for good reason: The technology has the potential to offer tremendous time- and cost-savings. However, the industry is still familiarizing itself with these guidelines and what they signal about where remote sensing stands today.
Power performance testing – past and present
When a project is underperforming, the cause is often not easy to identify. Is it the wind resource or the turbine itself, or was there a mistake in the initial energy assessment? Finding the culprit and evidence to prove it is a challenging and expensive process. Meanwhile, project owners are under substantial pressure to explain the situation to their investors and board and bring the project in line with revenue expectations.
Before the new IEC standard was released, if the problem was with the turbine, there was only one way to prove it. With warranty compensation at stake, a full IEC power performance test specified that the owner must follow the rules of its turbine supply agreement, hire a third-party analyst and install an expensive hub-height met tower to collect the measurements for the study.
Due to cost and time, it is a fairly routine practice to first conduct an informal (also called indicative) power performance test using a stand-alone remote sensing device to more quickly and easily investigate the issue and see if there is a simple fix to the problem. Using remote sensors in this way will continue to be commonplace in the industry due to the low costs and practicality.
Under the new standard, in cases of warranty compensation, a full IEC test is still required, which, again, must follow turbine supply agreements and involve a third party. However, now owners have several options for collecting measurements at relevant heights. Along with met towers at or above hub height, they can choose to use a remote sensor in conjunction with a hub-height or short met tower.
In this case, the remote sensor can be used to provide shear validation and other supplemental information or to provide all of the meaningful data for the power performance test. This case is the most compelling because of the cost-savings related to using a short met tower with a remote sensor and also because of the trust the IEC is placing on the remote sensing data for this highly complex and sensitive procedure.
It is also important to highlight what the guidelines do not cover and the specific limitations placed on the technology.
As mentioned above, the new standard does not allow the use of stand-alone remote sensors for this application. In all cases, a met tower is needed to provide a traceable path back to a controlled environment, specifically a wind tunnel, where met tower instruments have been calibrated.
The new standard also applies only to ground-based SoDAR and LiDAR and only in flat terrain, specifically prohibiting the use of remote sensors for this application in complex terrain. This is because remote sensing data often needs some correction at these locations – e.g., using CFD (computational fluid dynamics) or another flow model. Currently, there is no agreed-upon best practice for corrections, so this is not addressed in the new guidelines.
The IEC and wind resource assessment uncertainty
It is crucial to note that IEC 61400-12- 1 is not designed for characterizing uncertainty in wind resource assessment. Because the standard is founded on having a traceable path back to an anemometer in a wind tunnel, it calculates remote sensor uncertainty in a cumulative fashion, with each point of uncertainty (including those related to the site, tower and instruments) being added on top of the remote sensor’s own uncertainty. By design, uncertainty values achieved according to the standard will never be lower than the uncertainty of an anemometer, and they are overall quite conservative.
For the purposes of wind resource assessment, it is important to obtain an unbiased estimate of uncertainty (one that is neither too conservative nor too optimistic). While it can be tempting to use the IEC standard, in formal energy assessments, the results are ultimately overestimated or they contain higher uncertainty values, which are detrimental to the industry. This does not benefit the developer or the project in the long term because the inevitable consequence of higher uncertainty means unfavorable financing terms.
The best way to calculate the true uncertainty of a remote sensor as an instrumentation class for resource assessment purposes is to conduct a comparison study evaluating the device against a collocated met tower at a number of geographically distributed sites. In 2015, Vaisala, in partnership with its customers, conducted a studying finding that in simple terrain, its Triton remote sensor had essentially the same uncertainty as a class 1 anemometer. Two later verification studies conducted by Ecofys achieved similar uncertainty values at two different locations following very similar practices to those outlined in IEC 61400-12- 1 Annex L.3.
Breaking down Annex L: IEC requirements
In Annex L, the new standard describes the mandatory procedures required prior to using a remote sensor in a formal power performance test. The IEC stipulates that 1) the remote sensor must have a classification (provided by the manufacturer), 2) the owner must verify it annually against a tall tower (though this can be done at a different site), 3) uncertainty is calculated according to the standard, and 4) a short tower reaching the lower blade tip must be present during each power performance test as a “sanity check.”
Step 1: Remote Sensor Classification
Classification aims to systematically measure the device’s sensitivity to conditions that may vary between the calibration site and the measurement site to uncover any bias or error in these conditions. Remote sensors must be classified in the field rather than in a wind tunnel, so this is not a perfect test; it is susceptible to site conditions, tower variability and differences between units, meaning there will always be some level of bias.
Manufacturer classification requires multiple three-month measurement campaigns to capture the seasonal cycle, completed with a minimum of two units at two sites. Ideally, this is completed over a longer period of time at more sites, considering more data achieves more meaningful results. Some sources of bias can be corrected with new software or firmware; however, once implemented, the classification step must be carried out again. While this can be a slow process, it ultimately improves results and measurement accuracy.
Step 2: Verification
Following classification, the project owner must verify its device annually against a tall tower measuring at three heights (including +/-25% of turbine hub height) for a minimum of 7.5 days, excluding data-filtering. This can be done at a different location from where the power performance tests will later be conducted, but gathering data beyond 7.5 days is recommended considering you must measure conditions to which the device has atmospheric sensitivity, as identified in the classification stage. The data is then binned and fit to a slanted line using a slope and offset method. Having hub-height measurements from a tall tower during verification gives you the option to later use calibration to reduce measurement bias in the next step.
Step 3: Uncertainty Calculation
Before the device is used at the power performance test site, you must calculate its total certainty based on each height and wind speed bin and then square it within each bin so larger errors have a bigger effect than smaller ones. All sources of uncertainty contribute to the total uncertainty, even those not related to the remote sensor itself, including uncertainty associated with the anemometer, tower shadow, site, and distance between the sensor and the tower.
Measurement bias is a key contributor, which again can be calibrated if it reduces the overall uncertainty. Measurement noise is also included, but it is reduced the more data you collect during verification.
Carefully completing and documenting each of these steps is to your benefit because it reduces the overall uncertainty of the ensuing power performance tests.
Progress for remote sensing
While the process outlined above requires time and care, the inclusion of remote sensing devices in the new IEC standard is a strong signal of trust in the technology and recognition of its ability to more quickly and efficiently complete power performance tests.
Yet, it is still early, and the industry is still trying to fully understand the standard; device classification and adoption by turbine manufacturers and operators will take time. Meanwhile, informal (or indicative) power performance tests conducted using stand-alone remote sensors are taking off fast, demonstrating the technology’s ability to more easily and cost-effectively diagnose and resolve underperformance issues.
Lee Alnes is global manager of energy measurement systems, Brian Alves is remote sensing product manager and Francesca Davidson is energy communications expert at Vaisala.
Photo: Vaisala’s Triton Wind Profiler, a ground-based remote sensor system
