Routine wind-turbine maintenance once required a wind technician to suit up and climb up-tower just to inspect what components might need upkeep or repair. Today, wind farms can offer owners a whole new way to approach operations and maintenance, thanks to condition monitoring and advanced prognostic systems.
Condition monitoring is an O&M tool that helps wind-farm owners and operators monitor the health of turbine components and related electrical systems. Its purpose is to predict maintenance issues so site operators can conduct repairs and replacements only when needed to avoid unnecessary and costly up-tower jobs.
Although the intent is to cut time and cost from O&M tasks, condition monitoring system (CMS) have become rather detailed in accumulating and analyzing data and, therefore, costly.
“Over the last few years a lot has changed in the field of CMS and data acquisition,” says Dr. John Coultate, Head of Engineering Development at Romax Technology. “For example, the idea of putting ‘intelligence’ in the data-acquisition box or in turbine sensors has proved unnecessary — this just drives up cost and the limitations on data storage and transmission.”
Dr. Coultate points out that advancements in embedded computing also means that high-performance systems can be deployed at much lower cost. “Most systems today rely on an older and high-cost approach to architecture. But we’re learning that by using less expensive sensors and electronics in each condition monitoring unit — and only one in each turbine — the cost of CMS comes way down,” he says. “This is particularly important for operational wind-farm owners looking to retrofit their monitoring systems because they often don’t have the budget for new units.”
Of course, cheaper is not always more efficient so Coultate and his team at Romax had to think outside the box. “The traditional approach needed a change. So we put all the complex mathematical algorithms in Cloud computing — rather than on expensive sensors and electronics — and reduced the condition-monitoring unit to a simpler, more robust piece of equipment.”
This lets a wind farm transmit data collected from every sensor on its turbines, wirelessly or by Ethernet, back to a single server for data processing. Essentially, this provides three benefits: cost-savings on sensor equipment, a better more extensive data-collection reach, and a more scientific approach to data analyses.
Sentient Science’s Stephen Steen has also noted the digital change in monitoring capabilities. He is the Head of Industrial Internet Solutions for the company. “Many wind-turbine operators have begun the move from condition-based monitoring systems, which include vibration sensors, to a material science-based prognostics approach,” he says. So where condition monitoring once gave turbine operators a heads’ up regarding potential equipment failures to minimize turbine downtime and related costs, advanced prognostic systems can predict failures before they occur and let turbine operators plan ahead.
“An initial fleet risk-assessment can now provide the failure rate data of each individual turbine in a fleet over the next 20-plus years.” Steen says this means asset managers are able to build multi-year budgets and maintenance forecasts to lower their O&M costs. “The move to digitalization can also offer a rolling five-year forecast into the predictive health of each asset at the major system and component level.”
The forward-looking approach allows “virtual testing” of each asset using a material science-based method that can provide savings beyond lower O&M costs. “It facilitates supply-chain management because wind-farm operators can better anticipate when and where to expect component replacements. This digital approach also lets an operator simulate the impact of different supplier components and how they will affect asset life,” he says.
This is no small feat. CMS with advanced prognostics is providing turbine owners with a better understanding of their asset health at a micro-structural level, which means a more accurate prediction of failure rates and more cost-efficient and effective O&M planning. As Steen explains, this deeper level of asset life understanding improves a wind-farm owner’s ability to negotiate warranty and insurance contracts and strategize multi-year business plans and initiatives.
“Digitalization provides the forward visibility needed to reduce the cost of energy by $10/MWh as it enables smart decision making and multi-year forecasts,” he says. “Building virtual models of each asset on a wind farm and monitoring how they perform under the operating conditions is where the future of wind energy is moving.”