Enhanced recovery

Four-Injector Polymer Pilot Expands to 80 New Injectors Using Modular Concept

The complete paper presents steps to accelerate enhanced oil recovery (EOR) in a Grimbeek field from a four-injector pilot to 80 new injectors in a rapid deployment.


The complete paper presents steps to accelerate enhanced oil recovery (EOR) in a Grimbeek field from a four-injector pilot to 80 new injectors in a rapid deployment. The authors focus on a simple modular strategy rather than iterating multiple possible engineering solutions to distribute polymer across the field.


The goals of the Grimbeek field-­development plan include development of the field with a capital expenditure of $4–5/bbl or less and an operating expense (OPEX) of $7–8/bbl or less. The development risk is minimized through the use of a mobile deployment based on the plug-in concept that uses existing waterflooding installations and a sweet-spot-based agile strategy. The authors plan for a surfactant pilot for a second rotating cycle across the field.

The polymer used in the pilot was Flopaam 3630S. The incremental reserves were predicted from simulations that represent an incremental recovery factor per stratigraphic sequence. Rather than using early integration of studies and anticipation of facilities definitions, the authors fast-tracked the integration of laboratory and reservoir/architecture data. Tasks performed during the evaluation study are discussed in detail in the complete paper.

Reservoir Model Construction. The Grimbeek field is in depths of 1000 to 1300 m in the El Trebol formation corresponding to a fluvial environment. Total oil reserves are estimated to have an approximately 40% recovery factor (base case) or 50% (upside case). These reservoirs are a complex of fluvial unconsolidated sands.

Reservoir quality is excellent, with multidarcy permeability and porosity up to 28%. The oil is viscous (120 cp at 60°C reservoir temperature). Between five and seven stratigraphic sequences are grouped into two to four zones in each well. The reservoir has favorable conditions (low salinity of 11,000 ppm, low hardness of approximately 200 ppm, and a low temperature).

The geological model was initially a merging of the three models, which allowed tracking of the fluvial channels across the three reservoirs. These reservoir models have many grid cells (more than 1.5 million active cells). Running the models separately saves time. The individual models were run to investigate the effect of many geological uncertainties.

The Grimbeek block has three reservoirs, each a standalone simulation model (Grimbeek II, Grimbeek Norte, and Grimbeek Norte II). These reservoirs are independent from a dynamic point of view. The history match of the waterflood suggests a relatively high value of vertical communication, a crucial element to benefit from crossflow during polymer flooding.

Recovery Factor and Reserves. Uncertainties in microscopic displacement efficiency are related mainly to residual oil saturation (varies from 23 to 25%). The parameter is well-defined from extensive coreflooding and is matched by more than 8 years of waterflooding using a particle-swarm-assisted history-matching technique.

Current Waterflooding Installations. The injection system consists of clustered injectors connected to satellite stations. Eight satellite stations are in the Grimbeek II reservoir, and five satellite stations are in Grimbeek Norte; five additional satellite stations are planned for Grimbeek Norte II.

Each satellite station can connect eight to 10 injectors. Injectors are vertical; few are deviated sidetracks using a single location because of topological complexity. All injectors feature commingled completions.

Injector Intervention for Well Conditioning Before Polymer Injection. A campaign was designed to condition each polymer injector, including changing tubing installation for coated standard steel and removing, if necessary, formation damage in each injection zone. A criterion of 20% maximum loss acceptance of the well-­injectivity index for a remedial treatment was established.

Inline Multiphase Flowmetering. The production system is geographically dispersed, with multiple collectors and batteries. More than six wells produce on the same loop. Determining when to move the skid to another satellite is critical to the presented strategy. This task requires a good simulation model, which needs accurate production and injection allocations and, therefore, accurate production monitoring. Such monitoring is also required to define optimal reservoir production and injection management that eventually can increase plateau duration and minimize production decline. Taking these factors into consideration, a production-monitoring plan has been designed that increases the number of multiphase flowmeters per production loop dramatically. Remote control capabilities can increase by a factor of three or four the number of injectors and producers that can be monitored by a single operator, including the skids that provide the polymer solution to these injectors.


As of 2019, total volume injected was 0.56 pore volumes (PV). A total of 0.26 PV has been injected in 12-hour water-alternating-polymer mode. The polymer injection velocity is 0.15 PV per year. The skid is moved when the polymer injection is 0.4 to 0.6 PV and the cluster has delivered production. At that point, a water or a surfactant/polymer pilot is resumed.

The polymer concentration at the producers is a key parameter for monitoring process efficiency and calculating average effective adsorption and retention at the scale of the simulation model. Polymer production increases after more than 24 months from the start of polymer injection in the confined well.

Injection so far has been around the fracture pressure, and no early polymer breakthrough because of fracture propagation has been observed.

Incremental Costs, Effect on Schedule, and Economic Evaluation. In addition to the mobile injection skids, the incremental costs involve injector interventions to change to coated-steel tubing. The effect on the water-­separation system caused by produced polymer has been almost negligible in the pilot. Some impairment in the separation system after 36 months of polymer injection was observed; this was mitigated by changing the demulsifier. However, a longer water/oil separation time is expected with higher-viscosity produced water with polymer (40 to 50% more time to settle compared with water only). However, for now, dosing a standard production demulsifier solves the separation problem.

Polymer-Injection Cycle and Slug Size to Maximum Oil Rate. The ­polymer-injection slug is cluster-based and varies between 0.4 and 0.6 PV at 2,500‑ppm concentration. The injection rate is 0.15 PV per year. With the presented strategy, recovery factor can be increased over that of waterflooding in the range of 11 to 23% of stock-tank oil originally in place per stratigraphic sequence. Injection-pattern choice and well type can be designed as the peak oil is reached quickly (approximately 0.24 to 0.35 PV). One could argue that stopping the injection at peak oil brings sizeable incremental oil with 17 months of injection. This is an additional benefit of the mobile strategy. Fig. 1 shows that optimal slug size is approximately 0.6 contacted pore volumes of polymer solution injected.

Fig. 1—(a) Polymer flooding incremental recovery factor over waterflooding by date. (b) Polymer flooding incremental recovery factor increase over waterflooding per PV of polymer solution injected.


Water-Cut Reduction and Water Conservation. Simulations ­indicated a 15 to 25% reduction in the water percentage in each cluster of producers affected by polymer injection. The polymer injection project saves 1.2 million bbl of water. This considerable savings is sometimes difficult to quantify in terms of OPEX, however.

First Phase of Polymer-Injection Response (Water-Cut Reversal After 3 Months of Injection). Simulations indicated that incremental oil would start to ramp up after 6 months of stable polymer injection at the viscosity target. After approximately 2 months of injection (adjusting the viscosity in the polymer skid took approximately 6 weeks), the recovery process entered the first phase of response in which water cut stabilizes and begins to reverse the trend.


The complete paper presents an accelerated strategy to first polymer. Reservoir simulation is a key tool not only to design the project but also to put into perspective the benefits of polymer flooding. The waterflooding case should be calculated using the full injection potential per injection zone. Typically, the surface facilities needed to inject and treat a fluid volume 10 to 15 times greater than that of the polymer-flooding case, representing a huge investment that brings only a 3 to 5% incremental recovery factor.

In a relatively short period, the project moved from the laboratory to commercial upscaling on the basis of a successful pilot of more than 16% incremental oil and with a utility factor of 1.73 kg/bbl. The effective polymer retention at the reservoir scale interpreted with a reservoir-simulation model of 20 m and 1 m vertical is low (less than 0.5 μg/g).

Secure water quality has been important for maintaining viscosity and conformance. A mobile water-treatment plant was added that feeds the polymer-injection units to secure water quality regardless of the quality provided by the main plant supplying the waterflooding project.

This work led to a significant investment in expanding the technology in fluvial, multilayer, medium- to low-net-to-gross reservoirs.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper IPTC 20285, “A Successful 18% STOOIP Four-Injector Polymer Pilot Expands to 80 New Injectors in 6 Years by Adopting a Modular Concept in Grimbeek Fluvial Reservoirs,” by Juan E. Juri, Ana Ruiz, and Viviana Serrano, YPF, prepared for the 2020 International Petroleum Technology Conference, Dharan, Saudi Arabia, 13–15 January. The paper has not been peer reviewed. Copyright 2020 International Petroleum Technology Conference. Reproduced by permission.