Data-Driven Technologies Accelerate Planning for Mature-Field Rejuvenation

Fig. 1—Operating income and break-even oil price for all producing wells.

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

This paper presents a data-driven technology and associated work flow for fast identification of field-development opportunities in mature oil fields that accelerates subsurface field-development planning and reduces the time requirement from months to weeks. This technology is ideally suited for large, complex oil fields with large data sets and has been used for brownfield rejuvenation, for asset evaluation during acquisition, and as an independent validation system within internal review programs for large oil companies.

Introduction

This approach does not require full-field static or dynamic modeling but retains a high degree of confidence by coupling a reliance on data and analytics with judgment of experienced subject-matter experts who quickly validate results. This methodology relies on automated or machine-assisted geological and engineering work flows. Because the forecasting algorithms use regression or neural-network techniques that train on historical data, this approach is not applicable to undeveloped or early-stage greenfields.

Methodology

This technology relies on rapid integration and quality control of key geological and well data, including wireline logs, production/injection by well, completions data, and well trajectories. Automated, machine-assisted, or neural-network-driven work flows enable quick geological mapping, estimation of flow contribution by stratigraphic flow unit, decline-curve analysis, single and multitank material-balance history matching, fractional-flow modeling, drainage-area estimation for all wells by flow unit, identification of pay behind pipe, identification of unswept/bypassed-oil pockets, production mapping using numerical tracers, and waterflood optimization.

Reservoir-Engineering Work Flows

The first step of this approach is rapid analysis of production, injection, and pressure data. Numerous standard work flows of reservoir engineering are automated or machine-assisted during this phase.

Performance Summary of Active Wells. Maps and trends of oil, gas, and water production and well vintage and grouping analyses by layer and block provide a quick, automated framing analysis of reservoir trends.

Offline-Well Analysis. Special emphasis is given to wells currently offline, which often present the quickest and easiest opportunities to raise production. The cause of each well’s shut-in is characterized on the basis of trends in water cut, gas/oil ratio, and pressure. These analyses will lead to specific opportunities to restore production from offline wells.

Decline-Curve Analysis. The accuracy of automated decline-curve analysis is improved through automated event detection, wherein trend breakpoints are detected mathematically.

Reservoir-Contact Analysis. Analytics and statistics describing reservoir contact are gathered for all wells over time by zone, accounting for well type and lift mechanism. This analysis combines data from completions, formation tops, and production/injection databases to characterize the completion contact on a layer basis automatically.

Flow-Unit Allocation. The flow-unit-­allocation module estimates the production and injection for each formation, in each well, for each month, allowing a granular perspective in performance by formation and by well, which is used in many other modules of this methodology. All relevant data sets that are available are incorporated for every well over the field’s entire history.

Pressure and Voidage Analysis. Trends in reservoir pressure are described by block and by layer, and this is coupled with a voidage-replacement analysis. These results are used to improve or optimize pressure-maintenance strategies and to identify undepleted zones, and they are integrated into the analytics of other modules of this methodology.

Stochastic Material Balance. On the basis of the possible range of each input, a single- and multitank material-balance history match is performed stochastically, providing a range of likely outcomes and forecasts

Fractional-Flow Analysis and Heterogeneity Index. The oil, gas, and water fractional flow for all wells is provided over time, with the estimated source of water breakthrough provided when available. This is based on each well’s monthly production, pressure/volume/temperature curves, and estimated reservoir pressure at each timestep. Wells then are grouped according to the heterogeneity index, with specific wellbore solutions listed for each group.

Sweep-Efficiency Estimation. Areal sweep, vertical sweep, and displacement efficiency are quantified by use of analytical correlations. Total sweep efficiency is estimated as a function of pore volumes of water injected or pore volumes swept through aquifer influx.

Automated Pay-Behind-Pipe Work Flow

The work flow to identify recompletion opportunities targeting pay behind pipe consists of three main steps.

Step 1: Geological Mapping. On the basis of petrophysical log data, geological maps are prepared for key rock properties, leading to pay identification. These maps are coupled with a spatial distribution of oil in place to form the mapping foundation for the pay-behind-pipe work flow.

Step 2: Behind-Pipe Opportunities. Behind-pipe opportunities refer to target intervals that contain unswept oil in existing wells where additional recompletions are required for production. The identification of behind-pipe opportunities is accomplished through a multistep work flow that integrates both geological and engineering data.

Step 3: Production Forecasting. This phase of the work flow forecasts production for missed net-pay opportunities. The predicted attribute is user-selected and can be either initial production (IP) or estimated ultimate recovery (EUR). Note that if IP is selected as the attribute to predict, an EUR is estimated using decline parameters of active neighborhood wells.

Once the missed-net-pay opportunities have been identified and production has been forecast, confidence levels are defined on the basis of recovering the production target predicted in the preceding phase. To estimate confidence, the predicted target is compared with the performance of analog wells. For every opportunity, the analog-well set is defined as the set of wells producing from the opportunity’s target zone and located within the spatial neighborhood and within the temporal neighborhood.

Waterflood Optimization

The methodology for rapid, automated waterflood optimization is based on a numerical-tracer technology describing the relationship between producers and injectors. The objectives are to recommend specific opportunities to improve waterflood performance by altering injection patterns, layer targets, injection rates, and production-well rates or targets and to plan waterflood expansions targeting new layers or areas. This reduced-physics, data-driven technology relies principally on well-production rates, well-injection rates, and geological properties to characterize the strength and efficiency of producer/injector connections.

To quantify the strength of each interwell connection, the stationary pressure equation is solved and the solution is post-processed using the tracer equation to estimate well-allocation factors. The efficiency of each connection is determined by an empirical fractional-flow model calibrated to production data at each connection. This information is then fed to an optimization engine that takes into account the connection parameters, the various existing operational constraints at the field and well level, and the general objective of the strategy to propose a rebalancing of the producer and injector. Target rates are defined that will help strengthen efficient connections and weaken connections associated with swept areas. Geological data are used to create a reservoir grid where the connectivity analysis will be applied, and operational data of each well are used to calculate the strength and efficiency of the connections between injectors and producers as well as aquifers.

Results

The subject of this case study is a large, mature, conventional oil field in North America, comprising five stacked reservoirs with numerous sublayers, which has been under waterflood for more than 60 years. The oilfield data set is large, because of a large well set (more than 800 producers and injectors) and a long producing history (more than 85 years). Following the conventional model-based field-development-plan (FDP) work flow, a multidisciplinary technical team required many months to prepare a new FDP with associated production forecasts. The data-driven work flows presented in this paper allowed a new subsurface FDP to be prepared within 3 weeks, with a detailed plan for field rejuvenation.

Well Operations. An economic model is prepared for all producing wells in the field, quantifying the profitability of each well at current commodity prices. This process is depicted is Fig. 1 above, which plots the monthly income [revenue minus operational expenditures (OPEX)] of each well on the left axis and the break-even oil price for each well on the right axis. According to this analysis, 13 wells currently producing may have been uneconomic at the price of $43/bbl. These wells have been recommended for testing and possible shut-in or cycling and were added to the list of available wells in the production-uplift and water-shutoff categories.

Production Uplift. Fifty-one active wells and 13 inactive wells were identified with opportunities to raise production through wellbore rejuvenation, usually through optimized artificial lift or improved stimulation practices.

Recompletion Targeting Pay Behind Pipe. Thirty-six opportunities for recompletion targeting missed net pay were high-graded. These 36 opportunities were determined during the vetting process to be of high confidence and were recommended for immediate implementation, although some of the opportunities required a break-even oil price higher than $43/bbl.

Waterflood Optimization and Expansion. Opportunities to optimize the current waterflood were identified, relying on recompletion of existing injectors to target new zones or conversion of offline wells to injection targeting new areas.

New Drill Locations. Ten new drill locations were identified as infills or stepouts targeting areas and zones indicated as unswept using a methodology closely related to the missed-net-pay technique. These locations were high-graded as having high confidence, although many were found to require a break-even oil price above $43/bbl.

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 185751, “Using Data-Driven Technologies To Accelerate the Field-Development-Planning Process for Mature-Field Rejuvenation,” by Jeremy B. Brown, SPE, Amir Salehi, SPE, Wassim Benhallam, SPE, and Sebastien F. Matringe, SPE, Quantum Reservoir Impact International, prepared for the 2017 SPE Western Regional Meeting, Bakersfield, California, USA, 23–27 April. The paper has not been peer reviewed.

Data-Driven Technologies Accelerate Planning for Mature-Field Rejuvenation

01 January 2018

Volume: 70 | Issue: 1

STAY CONNECTED

Don't miss the latest content delivered to your email box weekly. Sign up for the JPT newsletter.