Mature fields

Achieving Well-Performance Optimization Through Work-Flow Automation

The Samarang oil field, offshore Sabah, Malaysia, is undergoing a redevelopment project with integrated operations. Several work flows were designed and deployed in order to achieve an early milestone of providing real-time well-performance monitoring, surveillance, and optimization.

Fig. 1—Data-quality funnel for Samarang IO data-driven work flows.
Source: SPE 173578

The Samarang field is located offshore Sabah, Malaysia. The field is undergoing a major redevelopment project with integrated operations (IO). In order to overcome a variety of challenges and to improve field awareness, several work flows were designed and deployed in order to achieve an early milestone of providing real-time well-performance monitoring, surveillance, and optimization. The paper discusses how these work flows were implemented in an integrated way to provide a modern decision-support system for the Samarang field.


Samarang is an old oil field and thus exhibits the characteristics of many mature fields, including declining production. It has been in production for more than 35 years from multiple, now aging, platforms with insufficient metering and monitoring processes. Given this situation, well status and uptime is unknown for many days, causing delays in mitigation and rectification of production issues. Most of the instrument measurements have been carried out on an ad hoc basis; various facility readings were measured manually, which subsequently deferred field review, causing loss of production. The main technique implemented for production optimization has been artificial lift, with more than 80% of wells and strings being gas lifted. It was necessary to find a better way to manage this production flow stream efficiently with an improved asset-management strategy.

Solution Overview

For Samarang, the IO solution is designed in such a way that work flows process data into actionable information, improve decision processes by leveraging technology, and equip people with systems and tools enabling enhanced decision support. An IO field creates value through enhanced asset management by focusing on decisions to improve reservoir drainage, production, and operations. As part of Samarang’s overall IO development strategy, the wells and facilities were equipped with the required instrumentation to support approximately 25 work flows.

Oil and gas asset activities and decisions span a range of time scales; therefore, the work flows are divided mainly into three main categories: fast-loop, medium-loop, and slow-loop decision cycles. These are then subdivided into seven different subdomain categories covering flow-assurance, well-performance, facilities-monitoring, artificial-lift, production-planning, enhanced-oil-recovery (EOR) -surveillance, and optimization work flows.

The complete paper covers the first two stages of IO implementation for Samarang, in which the first five work flows are designed, developed, tested, and implemented. The work flows are selected in a way that involves real-time, data-driven well-level decision processes with fast-loop decision cycles, comprising activities with a decision span of 0–1 days (e.g., well status and uptime, well-test validation, and rate estimation). The IO implementation also covers data- and model-driven work flows in the optimization and artificial-lift domain that cover medium-loop decision cycles typically involving activities with a decision span of 1–90 days (e.g., gas-lift surveillance, diagnostics, and optimization).

Solution Implementation

For the data-driven work flows, the most critical process is to ensure the quality, reliability, and accuracy of data. Real-time, high-frequency well and facility data are automatically transmitted from a remote-terminal unit (RTU) to data-history systems and all the way up to the production-platform database. Per the “garbage-in, garbage-out” concept, for automating the process and logic, process-unintended, even nonsensical, input data (“garbage in”) can produce undesired, often nonsensical, output (“garbage out”). Therefore, it was deemed necessary to implement a thorough quality-check process for data acquisition, aggregation, and validation. The data-quality funnel is depicted in Fig. 1 (above).

Because data are the foundation of these well-performance work flows, to ensure data quality, data are traced directly from their sources. Data tie-in and commissioning begins after ­instrumentation-testing, precommissioning, startup, and hand-over processes are completed successfully. Once transmitters are installed, calibration and testing are performed for all measurements by filling up the required check sheets. RTU and data-historian site-acceptance tests are performed, and the first level of data aggregation and quality check is performed. During later stages, some basic data-quality rules and logic are applied to the raw data and they are validated and approved. For this, a point-to-point check procedure was applied as a part of the commissioning process before aggregating the values for the engineering work flows. The confidence level of data quality improves from 20 to 100% as data go through the quality funnel. Swim-lane diagrams, together with responsibility-assignment-matrix charts, are used for clarity in defining role and responsibility.

The data-acquisition process is automated and is executed before the relevant work flows are executed. The data here are not only referred to as the physical-data types or parameters, but are also used later as information, after going through the work-flow cycle, and are finally used as knowledge for the end users. Because the work flows are interconnected and dependent on one other, the data are processed and flow through the series of work flows in the following sequence: well status, well-test validation, rate estimation, gas lift surveillance, diagnostics and optimization, and operational back allocation (these work flows are described in detail in the complete paper).

Work-Flow Application and Results

The work flows are implemented at the Samarang Operations office in Kota Kinabalu (KK). The main access to these work flows is from the “Samarang Smart View” screens as a common asset-decision-support system shared by everyone.

By having fully integrated work flows in Samarang, engineers are able to acknowledge exact well status and accurate uptime continuously and perform mitigation actions proactively. This also assists in managing unplanned events and prevents production and injection deferment by improving well uptime. Production trends are estimated for active wells instantly when there are changes in operating conditions. Estimated production is also used in operational back allocation and reservoir modeling, leading to better field management.

Because most of the wells in Samarang are gas lifted, continuous real-time surveillance of gas lift systems provides an excellent insight to determine if the well is operating close to optimum conditions. The entire process is integrated and automated, from data collection to final outputs of visualization, allowing management by exception by means of warning and alarm notification. The work flows are interrelated and integrated, such that the results and key performance indicators (KPIs) of each work flow are analyzed in a way that allows engineers from different domains to collaborate for better field management.

Work-Flow Operational Guidelines. Operational guidelines (OGs) are a result of the process of work-flow operation after the work flow is implemented and commissioned. These guidelines describe the details of various IO work flows, which include who will be involved, how the work flow will be used in day-to-day operations, and how it can be useful for operational decision making.

These processes were implemented by using visual representations of swim-lane diagrams. OGs act as a guide to illustrate the use of the “To-Be” work flow by the actors in day-to-day operations and do not represent the IO system logic but rather the logic of how people should be using the work flow. This includes who should be involved (actors) for each phase of the work-flow life cycle; the roles, guidelines, and boundaries of each actor; targeted Smart View screens for analysis; and how the work flow can be used more effectively.

Success Cases

This section describes two successful cases (more are provided in the complete paper) that have been achieved through successful data- and ­model-driven automated work-flow implementation that allows faster, effective, and collaborative diagnostic decision making and field implementation to gain value.

Success Case 1: A Collaborative Working Environment (CWE). During daily video conferencing between the KK staff in Samarang and headquarters in Kuala Lumpur, engineers are able to discuss any issue pertaining to the wells and to create solutions. The Production Surveillance team is able to analyze well performance through gas lift optimization even at remote jackets. The CWE has enabled the team to have effective meetings by having real-time work-flow KPI visualizations and reliable audio/video conferencing.

Success Case 2: Enhanced Gas Lift Diagnostics. The value of a gas lift diagnostics and optimization work flow mainly comes from automating most of the processes during this work flow that were previously handled manually. In this particular case, the gas lift diagnostic work flow raised an alarm that the well was multipointing. On the basis of well-test parameters and further gas lift diagnostics, multipointing was confirmed. This was an opportunity to improve current well-production performance. Engineers executed the work flow and performed detailed diagnostics to troubleshoot the problem. KK engineers worked as a team and, by running sensitivity studies on operating conditions, achieved the single deepest injection point. Further diagnosis identified that the casing-head pressure (CHP) was too high. The solution was to ensure the CHP to be approximately 650 psi for optimum injection depth. Production technologists in KK advised offshore personnel to reduce the CHP to optimize the well production. This has led to reduced gas lift consumption, from 0.9 to 0.4 MMscf/D, owing to the achievement of single-point injection. By optimizing gas lift, the potential gain is approximately 200 B/D.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 173578, “Samarang Integrated Operations: Achieving Well-Performance Monitoring, Surveillance, and Optimization Through Data- and Model-Driven Work-Flow Automation,” by M. Zul Izzi Ahmad and Colinus Lajim Sayung, Petronas, and Muzahidin M. Salim, M. Kasim Som, Lee Hin Wong, Shripad Biniwale, Nur Erziyati, Kenneth Soh, Roland Hermann, Vo Tri Nghia, Lau Chong Ee, and Muhammad Firdaus Hassan, Schlumberger, prepared for the 2015 SPE Digital Energy Conference and Exhibition, The Woodlands, Texas, USA, 3–5 March. The paper has not been peer reviewed.