A leading International utility services company faced shrinking revenues, leading them to turn to modern digital technology. By using the AF Accelerator program, the team increased their digital expertise and tooling, providing a way to expand the reach of their RPM center and add new revenue in adjacent markets. With Element, they were able to better utilize production data, improve their competitiveness, increase profitability, improve compliance with PPAs, reduce their exposure to the spot market, and recognize new revenues through excess energy sales.
The global increase in the use of relatively clean-burning natural gas has restructured global gas markets. For existing providers in the natural gas services business, this has meant operating in an increasingly competitive — and regulated — market. As a consequence, this international utility services company’s revenues shrank, leading them to use the latest digital technologies to better utilize production data. Much of the center’s infrastructure was built around legacy applications housing inflexible dashboards that were difficult to customize and adapt to reflect new data sources, slowing their decision making. This, coupled with the desire to take advantage of the latest technologies, led them to a new, digital approach.
Their first digitization project was targeted at its Reliability Performance Monitoring (RPM) center that monitors its fleet of combined-cycle natural gas turbines totaling over 4 GW of capacity. The center was established 5 years earlier as a comprehensive asset management platform for 4 GW of operating capacity across their 8 sites. The existing RPM Center was focused in two areas: 1. Condition Based Monitoring (CBM) to improve turbine reliability, reduce operating & maintenance (O&M) costs, and increase Asset Utilization, and 2. Performance Monitoring to increase production while reducing fuel consumption. The center wanted to leverage OSIsoft PI Asset Frameworks (AFs), time-series, and transactional maintenance records from SAP to increase net capacity by maximizing turbine utilization. This required improving turbine reliability by using more effective maintenance strategies to reduce unscheduled forced outages.
The team chose OSIsoft PI System as the core of the new system, providing a standardized, single source of truth to store plant information, including DCS and historian data, predictive maintenance results, and day ahead market analytics.
Building the AF took considerable time and effort. Each site engineer needed to assemble maintenance data from SAP with over 10k tags/plant from OSIsoft PI (measuring pump vibration and temperature). Using spreadsheets proved ineffective as it not only introduced errors but was slow, impacting their ability to monitor turbine efficiency.
And because new data was manually assembled within spreadsheets, KPIs that tracked excess energy were built site-by-site leveraging those same spreadsheets to display scheduled and unscheduled availability. This made comparing KPIs for like-units between sites difficult - which in turn - slowed access to turbine performance data, and impacting contractual production vs potential capacity decision making.Maintenance also suffered as crews did not have enough time — nor insight — to perform corrective maintenance tasks efficiently. An investigation by the monitoring center’s data analyst also showed that data quality was an issue — especially with null and flat-lined values. This resulted in day ahead market analytics incorrectly forecasting demand leading to increased spot trading. Laborious data cleansing tasks by the PI admin and data managers to address the quality issue further impacted the team's ability to respond quickly.
Element delivered its Asset Accelerator program, which is based on Element AssetHub, and was deployed by AF experts using a proven methodology. AssetHub was hosted on Element’s Azure tenant and connected to OSIsoft PI server data archives. By contextualizing this data with maintenance records from SAP, AssetHub created Asset Twins of the gas plant and turbines. Data was then constructed as hierarchies and shared as PI AFs with PI ProcessBook to support new dashboards and KPIs.
With Element AssetHub, data is shared as AFs that drive real-time monitoring dashboards and KPIs within PI ProcessBook.
In addition, the team experienced:
Operators and the maintenance team have a “second set of eyes” as historical maintenance records are contextualized alongside time-series data across all sites using AFs. Dashboards provide earlier warning of equipment issues, helping to prioritize work and organizing resources and maintenance/crews in a timely fashion, increasing uptime and asset utilization.
KPIs roll-up new data automatically and span multiple sites. This removes the burdensome hand-coding of individual site spreadsheets and provides a way for engineers to compare like-unit performance to resolve issues consistently across sites. AssetHub builds — and automatically updates — AFs to reflect changes in production data, allowing engineers to focus on improvement activities without additional resources, increasing their efficiency 4x, and reducing costs.
AssetHub lowered risks and unlocked higher production by increasing compliance with PPAs, reducing exposure to spot markets, and materializing excess energy sales. Data is now trusted as Data Quality reports identify issues before analysis, reducing the risk of inaccurate predictions and ineffective maintenance tasks, improving asset lifecycle management.
The AF Accelerator program improved the efficiency of building AFs by 4x, reducing our O&M costs. Overall data quality also improved, minimizing the risk of incorrect day-ahead analysis. By helping to improve our maintenance prioritization, we increased turbine uptime and revenues.
By using the AF Accelerator program, the team increased their digital expertise and tooling, providing a way to expand the reach of the RPM center and add new revenue in adjacent markets.
I want to know where I have bad data before performing analysis as we waste so much time on work we cannot use.