Producing smart fields, now

Continuously knowing what a company’s wells and reservoirs are producing facilitates improved asset management and integrity, argues Shell’s Ron Cramer.

“Shell’s challenge now is to scale up these benefits to full global brown-field and green-field operations and to transform the traditional manual operations culture into a new ‘Smart Fields’ way of working”
-Ron Cramer

With the vast majority of industry wells lacking a continuous, reliable measure of well/reservoir performance, a key issue in upstream E&P operations is how to manage wells and reservoirs more effectively. Yet, if we cannot measure continuously, how can we manage better? How do we monitor well and reservoir performance? How do we perform hydrocarbon accounting? How do we report production for our wells?

The industry has traditionally used discontinuous well testing to determine well performance, as wells are tested once per month and it is assumed that for the other 29 days the wells produce the same. Mother Nature, however, is usually not that predictable. Also, the quality and accuracy of well tests are often unsatisfactory, so some tests are rejected and have to be repeated. At the heart of this conundrum is multi-phase flow. Almost no wells deliver a clean, measurable, single-phase stream, and it is impractical to install multi-phase flow meters or test separators on all wells.

Shell’s FieldWare Production Universe (PU) is a software application that continuously estimates oil, gas and water flows for all of an operation’s wells, all of the time. PU enables improved well surveillance, more accurate hydrocarbon accounting, automatic production reporting and production optimization. It safeguards the technical integrity of wells and reservoirs (for example, it provides early detection and control of gas or water breakout). And it is cost-effective in that it requires minimal commodity instrumentation and IT systems, much of which may already be present in field operations. In some respects, the technology provides such a step change that successful PU deployment often requires changing the way one operates and motivating the people involved.

How does it work?
PU uses dynamic data-driven models of the production system. The well models estimate water, oil and gas production flows in real-time, primarily from existing well instrumentation. Effects such as backing-out of weaker producers at headers are captured in these models. Physical models are not used – no well tubing diameters, no roughness, no fluid properties, no near well bore ‘skins’, and no pre-assumed multiphase flow correlations. Real-time well measurements are related to volumetric flows from test separators. The data-driven approach has been proven to be robust, usable and sustainable in the oil and gas production environment.

A key aspect is the Deliberately Disturbed Well Test (DDWT), which is used to characterise well performance. These tests go beyond traditional production well testing. The objective is to relate well production (oil, gas, water) to simultaneously measured well parameters, such as flowing tubing head pressure, gas-lift rate, etc. The emphasis is on capturing the response of the well to step changes in controllable parameters.

Once created, the individual well models are used to compute the well production per stream. PU accumulates daily flow per well, which reflects the actual producing conditions, including trips and restarts and plant operating mode changes.

A simplified abstract topography is constructed relating wells to a calibration point. Typically, the calibration point is a bulk separator continuously providing oil, water and gas measurements from wells in a given production system.

PU production data per well are compared and reconciled automatically against the installation’s overall export meter. This provides a reconciliation factor for each produced/injected stream on a continuous basis for the current day and the last 24 hours. Also in this graphical user interface is a diagnostics panel that alerts the user to production systems events. Event detection can be single point measurements or a complex logical mask to detect a specific event (for instance, contamination of the water disposal stream with oil). There is also an information panel, which alerts defective instruments and communications infrastructure.

With this single screen, an asset manager can gauge the current health of the production systems. If all the reconciliation factors are within acceptable bounds, then the production system is under control. If this is not the case, it is possible to drill down to process, header and well-level.

The output from the measurements on the bulk separator provide a 24/7 data stream at one-minute or more frequent intervals. PU uses the dynamic variation seen at the calibration point to further tune its well models. Plant trips and restarts are very visible and generate a lot of useful data, especially when the field is brought back online. The dynamic well models are updated every 24 hours to reflect the total information available in the preceding period, allowing tracking of decline in well rate, and increase in gas oil ratio (GOR) and water cut.

PU thrives on dynamics (for example, well bean up/bean down) to continuously update individual well models. Normal E&P operations provide a dynamic environment with well interventions, process trips, etc. If assets exhibit stable production with minimal dynamics, then dynamics can be introduced. Wells can be beaned up/down for short periods to cause transients to ripple through the process. Single or multiple disturbances can be introduced simultaneously. These pseudo-tests are known as Deliberately Disturbed Production Tests (DDPTs). If these tests are insufficient to re-align the models, then PU initiates a full DDWT.

PU is currently running on more than 1500 wells in 14 Shell operating units (OU) and affiliates, covering about 35 percent of Shell’s global production.

What’s coming down the PU pipe?
Real-time estimates of oil, gas and water flows for all wells is valuable for surveillance purposes. A logical next step is to use that information for real-time optimisation of reservoir, wells and the production process. PU-based real-time optimisation has been piloted for gas-lift optimisation and is being developed for beam pumps.

For example, PU Real-time Optimisation (PU RTO) is currently operational on an offshore gas lifted platform. It incorporates the basic PU functionality extended to include an optimisation algorithm for adjusting lift gas injected to each well for maximum production using minimal lift gas. PU RTO continuously computes optimal set points based on an inbuilt model and estimates of the amount of oil, gas and water that each well is producing; set points are automatically transmitted to the well gas-lift injection valves via the platform control system. The system controls eight wells to optimise the overall gross production.

The PU optimisation was installed some two years after the basic surveillance module, which had already demonstrated significant surveillance/optimisation gains (production gain of 370m3/day, increased well potential of 600-900m3/d and 20 percent reduction in utilisation of platform lift gas).

The subsequent PU RTO deployment has stabilised gas lifted production for this field. Individual wells have been optimised effectively and PU RTO has demonstrated its ability to rapidly detect dead wells subsequent to a well or platform trip (re-instating 670m3/d potential). It also has been demonstrated that sub-optimal lift settings can decrease gross production by more than 10 percent.

The new system allowed header interaction effects on well production to be quantified. After experiments were completed, a new production line was installed in June 2004. Installation of this line provided a gain of 11 percent in platform gross production, and also generated a number of non-quantifiable benefits. For instance, changes in well performance are noticed quickly and countermeasures initiated. PU automatically notifies staff of well performance and of significant changes via e-mail. In addition, since PU installation the performance of wells and instrumentation is highly transparent and the level of attention given to the facility has much increased. Unscheduled deferment due to process problems is now significantly lower as compared to other similar platforms in the operating unit.

As a result, PU is now being rolled-out across all Shell upstream assets. Each candidate field is assessed carefully in a readiness check to identify what repairs are required to existing instrumentation and, if required, where additional instruments need to be added before installation will start.

Conclusions
PU is well established in multiple Shell OUs, sufficient to establish significant benefits for more than 30 projects completed to-date: a five to 20 percent increase in production due to improved surveillance and optimization; up to five percent reduced operating costs due to optimisation (for example, reduced gas-lift gas and logistics savings due to reduced travel to the wells); and safer operations due to reduced operator exposure to hazard.

Shell’s challenge now is to scale up these benefits to full global brown-field and green-field operations and to transform the traditional manual operations culture into a new ‘Smart Fields’ way of working based on remote surveillance and control. Good progress is being made along this road, with some 60 projects in global assets planned over the next two years.

Ron Cramer works in the area of oil field automation and production systems and is a Senior Advisor for Advanced Production Management with Shell Global Solutions in Houston. He has 30 years’ experience with Shell International E&P in upstream oil field operations and production systems.

Shell experiences with PU
Formal PU post implementation reviews (PIR) have been conducted in three Shell OUs.

Operating Unit 1
PU was initially installed in June 2002. The objectives of the implementation were to test the technology in an operational environment and to document business benefits.

The field consists of 15 wells producing from three reservoirs. The three reservoirs consist of a mixture of free flowing, gas lifted and dual completion wells. The field is well operated, instrumented and maintained with >99 percent availability. Production data stem from the local DCS system.

Field operational strategy is to maximise oil production within the constraints of the gas export to the local domestic gas company, which is achieved by operating wells with low gas oil ratio (GOR). A key requirement to achieve this strategy is knowing the relative composition of well outputs. Before PU implementation, wells were tested monthly and the test results used for the following month’s optimisation, along with other calculations, such as deferment values. Well GOR changes frequently and PU quantifies these changes as they happen, facilitating continuous optimization, whereas prior to PU implementation they were operating ‘blind’.

Following PU implementation, the annual decline in production rate has decreased from approximately 20 percent to seven percent, and large monthly swings in oil production have been reduced, increasing the confidence with which production forecasts can be made. Total deferment has also been reduced by 2.2 percent. Production has increased by approximately 30 percent when compared to forecasts made before the introduction of PU. Other benefits include:

• Reduced well testing. The test separator was unusable for 15 months, and PU was the only method available for well surveillance.
• A marked increase in the stability of monthly oil production – 50 percent reduction in monthly standard deviation of values.
• Despite a 15 percent reduction in gas exports, oil production has met or exceeded targets since introduction of PU.
• More accurate testing of dual completed gas lifted wells.
• Instantaneous detection of process events. PU surveillance has enabled a number of events, such as water leaking into an export stream, to be quickly detected and remedial action taken.
• Operation with 50 percent reduction in staff levels. PU enabled the production programmer to continue production system optimization with increased overall workload from other fields whilst reducing staff levels.

Operating Unit 2
PU was implemented in February 2005. PIR findings indicate a three to five percent production gain due to real-time production surveillance due to faster well-fault identification/correction. Other benefits include:

• Opex benefit of $750,000 per year (one boat and two less positions).
• FieldWare PU has reduced the need for intra-field travel and thus reduced HSE exposure.
• Minimizing of hidden deferment – more accurate deferment reporting. In the first two months of 2005, a total of 14,000 bbl were reallocated (0.5 percent).
• Reduction of well-test frequency due to PU calibration/maintenance to a level whereby decline behavior can be modeled within PU.
• Improved surface process surveillance – PU flagged unstable station operation due to process control instability.

The PIR team concluded that a sustainable installation of FieldWare PU has been achieved in the Operating Unit and recommended implementation of the PU Real-Time Optimisation Module.

Operating Unit 3
The PIR covered readiness check, implementation and initial operation of PU on a field producing 7000 boe/day with 10 electrical submersible pumps. Following readiness checks, the PU implementation project started in January 2005. The following was concluded:

• The daily deferment volumes automatically calculated by PU are more accurate – 30 percent differences in deferment reporting were noted.
• PU can help unlock hidden deferment using accurate real-time rates for well oil, gas and water flows.
• PU has flagged multi-phase flow-meter calibration problems.
• PU is effective for real-time monitoring, as it provides timely information to those who need to know.
• Above benefits have led to a decision to roll out FieldWare PU to all Shell assets over the next three years.