Case Study Toward Digital Oil Field: How the ESP Operation is Changing by Using Automatic Well Models in PDO's ESP Fields
Authors
Atika Bimani (Petroleum Development Oman) | Rahul Kulkarni (Petroleum Development Oman) | Chia Yee Lee (Petroleum Development Oman) | Valerio Giuliani (Petroleum Development Oman) | Maryam Musallami (Petroleum Development Oman) | Abdullah Awaid (Petroleum Development Oman) | Nada Kamyani (Petroleum Development Oman) | Nitish Kumar (Weatherford) | Rahul Gala (Weatherford)
Publisher
SPE - Society of Petroleum Engineers
Publication Date
May 13, 2019
Source
SPE Gulf Coast Section Electric Submersible Pumps Symposium, 13-17 May, The Woodlands, Texas, USA
Paper ID
SPE-194414-MS
Abstract
This paper presents a case study on how Petroleum Development Oman (PDO) is advancing Electrical Submersible Pump (ESP) operations by using real-time Automated Well Model and delivering tangible values in ESP life cycle from design to surveillance and optimization. The value to be discussed is captured from actual examples found from all ESP wells in operation in PDO.
PDO's Well Management System (WMS) contains 2,000 real-time well models sampled at five minute intervals automating ESP well analysis such as ESP design, performance, well test validation and optimization, amongst others. This was achieved by aligning diverse data feeds with relevant business validation rules to auto-generate Online Well Models, once ESP is commissioned. To ensure the end users’ competence and sustainability of WMS usage, Standard Operating Procedures (SOPs) have been established for all activities such as ESP design, monitoring performance, well test validation, and intelligent surveillance by exception (EBS). Furthermore, this WMS is embedded in PDO's in-house ESP training to Production and Operation Engineers, where activity functionalities such as ESP design, performance monitoring, optimization and sureveillance by exceptions are routinely practiced in the class using real-time well data.
In the last 3½ years, PDO has successfully utilized WMS with tangible results. This case study details real examples where production and operation engineers have utilized real-time well models on the following functionalities shown in Fig. 1:
ESP Design: 30% of total submitted designs by vendors, which have been technically quality checked, needed to be revised on the area of production, pump setting depth, pump size, motor capacity and applicability of variable speed drive for optimum operation, increased efficiency and cost saving. This exercise is expected to increase ESP run-life thereby reducing the Company's intervention cost while minimizing deferred oil.
ESP Performance: All newly commissioned wells are instantly evaluated as per design targets. Since the start of utilization of this customized WMS, 1,500 new well tests have been validated by engineers, which ensured correct oil allocation for around 44% of total ESP wells across PDO.
ESP Optimization: Low hanging opportunities, i.e. by simply increasing pump frequency or reducing tubing head pressure (THP), have been capitalized on, which yielded nearly 2% increase of net oil production of ESPs at zero cost.
Pattern Recognition Exception Based Surveillance (EBS): Pattern Recognition EBS enabled timely planning for well intervention and reducing deferment by 30%.
Equipment/Pump operational limits: Around 50% of ESP wells operating in upthrust or downthrust were quickly observed and were successfully brought within recommended optimum range, improving ESP run life and reducing power consumption.
Additional functionalities and automation: Easily available real-time calculated variables such as pump efficiency, system efficiency, power consumption and inflow rate have eased the asset well review tasks, enhancing decision making and improving the overall ESP key performance indicator to over 85%.
The use of automatic well models ensures that at all times, engineering equations are used for decision making throughout the ESP life cycle. In addition, the well model based well test validation process brings the notion of virtual metering which ultimately will reduce the frequency of well test activities and reduce the amount of deferred production. Moreover, well models integrated with control algorithms have a wider scope of application on well on/off, water injection response/efficiency and station capacity optimization decisions