DCP Midstream: Achieving a rapid business transformation with the PI System
For over 90 years, DCP Midstream has provided natural gas gathering, processing, and transportation services for its upstream O&G customers. Collecting and processing 12 percent of the country’s gas supply, DCP Midstream is also one of the largest producers of NGL. In today’s commodity and price driven market, the company differentiates itself with digitally enabled operational excellence and delivers a quality product to its customers efficiently and reliably, at an affordable price.
To stay competitive in an ever changing market, in 2015, DCP Midstream implemented “DCP 2020”, an operational excellence program focused on business sustainability through efficiency, reliability and risk management.
Ambitious goals equals a rapid implementation
Meeting its DCP 2020 goals was no small feat. The team lacked the right visibility into DCP’s 60 gas plants, 9 fractionation plants, 57,000 miles of gathering pipeline, nearly 400 booster stations, 1400 compression units and 4,500 miles of natural gas pipelines. Existing systems were disaggregated and geared toward control and operations, not reporting and analytics. Crews needed a single, contextualized view of all operations as well as real-time insights, and they needed it quickly. The company embarked on an ambitious digital transformation journey leveraging the PI System.
In December 2016, with no prior experience with the PI System, DCP Midstream signed an Enterprise Agreement (EA) with OSIsoft and installed the PI System data infrastructure across the entire enterprise in less than two months. At the same time, the company designed and unrolled its Integrated Collaboration Center (ICC) at its headquarters in Denver, CO. The ICC centralized operational insights and enabled coordinated real-time decision making across the organization.
To accelerate the development of digital applications and solutions, the company created an Energy Lab team using agile development process. The teams leveraged Asset Framework (AF) to create smart asset object templates for real-time analytics, alerts, and notifications. DCP soon had over 580,000 tags, 8,200 AF elements and 320 PI Vision displays, most of which were created by the end users without any additional programming efforts.
The result was a hierarchical view of real-time operational data coupled with contextual data that crews could easily visualized to make data-driven decisions.
A plant based view
While the ICC monitored operational performance and provided critical feedback, real-time data was also made available on the plant floor. But DCP Midstream was not content to stop there. For example, gas plant operations depend on a number of compositions, making it difficult to optimize. Now, the company could visualize real-time plant operations for things like pressure and temperature, and couple it with simulated optimal performance metrics. The blending of data enabled operators to see any deviations in real time while understanding how actual performance varied from set targets. In addition, operational and simulated data was linked to financial information, assigning a monetary value to any changes that an operator would make.
Not only could operators now see how to make real changes that impacted the bottom line, DCP Midstream setup notifications to let them know when an asset is operating more than 10 percent off of optimal. This enabled operators to understand the root cause and submit reason codes for the downtime. With this type of transparency, operators became key stakeholders all working towards a common goal.
Isolation issues in a maze of pipeline
In the company’s maze of pipelines, there are over 1,500 compressor and pumps that gather natural gas from the well heads and transport it to plants for processing and NGL extraction, and then move it to market in large transmission pipelines. The compressors increase the pressure of gas so it can travel to its intended destination. These compressor and pump booster stations are located throughout the natural gas and NGL pipelines, and it’s extremely important that they are running efficiently and reliably.
To keep booster stations running, operators make rounds to visually inspect the stations. Field crews typically start in the early morning hours and travel for miles, which means they don’t inspect some boosters until later in the afternoon. However, after the PI System was installed with real-time analytics on their compressor fleet, one operator awoke to an email notification about a change in the plan versus actual compressor performance. After looking at the data in the PI System, he went to the site right away and was able to quickly fix the problem. Thanks to the notification, the issue was resolved first thing, rather than the malfunction persisting throughout the day with a potential failure. Today, contextualized, real-time operational data and analytics provided by the PI System and Energy Lab digital applications are changing how DCP Midstream operates. Using predictive analytics in Asset Framework, the company is moving away from its traditional maintenance plans in favor of condition-based and predictive maintenance programs. And it’s all paying off. In the first year alone, DCP saved $20-25 million thanks to improved gas and NGL plant operation, asset reliability and ICC coordination, effectively balancing out its initial investment. With the Enterprise Agreement, DCP has the ability to scale its data infrastructure without any licensing headaches and barriers to growth, and the future is looking bright.
Natural gas for the future
While the first year return on the PI System was stellar, DCP Midstream’s digital transformation certainly isn’t done. In the future, they plan to link operational data with geographical data and geospatial maps as popups within Asset Framework so they can optimize gas routing across 57,000 miles of gathering pipelines. In addition, DCP Midstream is exploring IIoT-enabled advanced machinery analytics to leverage more detailed collection of data from compression units as well as a smart booster stations, enabled by real-time data displayed within virtual reality (VR) headsets.