July 5, 2018, by Azeez Mohammed, President and CEO, GE’s Power Conversion
(Renewable Energy World)
As the solar industry continues on its quest to achieve lower and lower costs, the next step is to use digitalization to achieve further cost reductions from the bottom up — reducing unplanned downtime, improving asset reliability and availability, and guaranteeing long-term plant performance.
Putting “Digital Twins” To Work
One technique that is helping cost-conscious solar developers is known as “digital twins.” This involves creating a virtual plant asset replica based on normal operating profile data that can then be used to simulate how a plant should perform under optimum working conditions. By comparing real-time asset performance with its perfectly healthy virtual equivalent, asset performance management (APM) can help fine-tune asset parameters to enable peak productivity.
But more importantly, it can flag up when something is wrong or needs fixing — alerting operators ahead of time to tackle potential operational headaches before they take hold. And it can also enable them to pre-order the correct part for “just-in-time” delivery and carry out services after sunset — so no valuable production is lost.
Plant maintenance programs have traditionally revolved around the calendar — for example stipulating component checks or replacement on a quarterly, twice-yearly or annual basis. The downside to this approach is that it can miss faults that occur between service intervals and/or increase costs in terms of money and time for maintenance that simply was not needed in the first place.
By applying a predictive philosophy — a successful and long-established practice in the aviation industry — maintenance is carried out only when it’s required, based on data-driven intelligence. This reduces downtime, lowers costs and maximizes revenues — with no compromise around performance effectiveness.
Dispatchable Solar Power
If digitalization has a major role to play at plant level, the implications for the broader solar ecosystem are even more wide-ranging. Digitalization has the potential to enable a true “smart grid,” which uses battery storage and the “smart” release of renewable-generated energy, as well as a more responsive service for end users. Digitalization makes it possible and easier to synchronize production with the weather forecast, better balance the power demand/consumption curve, and let the storage system itself decide how much power to store — and when. It also means operators can maximize output and potentially avoid the need for any additional fossil fuel-based power generation when the sun dips behind the clouds.
Field Tests Underway
In late 2017, we started working with Invenergy — North America’s largest, independent, privately-held renewable energy company. Initially, we will deploy GE’s solar APM software, powered by Predix, on a 20-megawatt (MW) solar farm. The software will monitor and run analytics on the seven GE inverters that transform the DC electricity generated by solar cells into the AC electricity people use every day at home and work.
Each inverter will be equipped with sensors — small, secure data-collection devices that will draw information from the inverters and feed it into the cloud. There, the Predix cloud platform will clean the data, analyze it, and turn it into a useful set of advisories — information then used to monitor 200 different datasets such as critical component temperatures and voltages, which will be measured every 15 minutes.
The partnership itself will be a learning process for both GE and our customer — with the first phase all about collecting the maximum amount of data. We will then trim it down to focus only on the information that’s critical for the data science. In other words, sifting through the massive data pool to arrive at “pure intelligence.”
Phase 2 will further evolve as the “digital twin” built for the Invenergy project will continue to acquire information from the field. We expect to learn further from the analytics results and from the more than 70,000 assets installed by GE globally. This data will not only shed new light on maintenance effectiveness, but it will also help us explore other possibilities to further improve efficiency.
The daily routine of plant managers will also be transformed. They will simply open up the relevant page on a digital device each morning to see the issues they’re likely to face and will need to deal with during the day. It’s a very similar approach to how you or I might go to the office and read through emails in the morning before categorizing problems and actions to set up the day’s agenda.
Once the plant is digitalized, our goal is to achieve more than 99 percent plant availability throughout the renewable farm’s lifespan. Improved plant availability could translate into reduced operation and maintenance costs — alongside increased production — equal to US $200,000 worth of additional value annually for a 20-MW solar farm.
Invenergy is not the only pioneer to be discovering the advantages of the digital solar farm. We are at work to deploy similar modules at other solar plants, with a combined total production capacity of 2 GW, together with our customers illustrating how digitalization is writing a brand-new chapter for our fast-growing industry.