JPT logo
SPE logo

Technology Update

Advanced Desander Removes Solids Without Deferring Well Production

As oil and gas fields mature, production of sand and other reservoir solids can become an increasing and sometimes major impediment to hydrocarbon production and facility operations. These solids can move through the wellbore into process facilities, causing flow assurance problems that affect system integrity, export line transportation, cleaning, and disposal. As operators seek to optimize production from mature assets, solids management is likely to take on additional importance. Sand and solids production can also be a problem on some new wells.

To address these issues, Stork Technical Services developed an advanced online desander (AOD) for managing separator solids, providing thermographic determination of sand accumulation, and enabling clean sand discharge to the sea, with no deferred production (Fig. 1). The system has been used successfully in eight deployments on the Hummingbird floating production, storage, and offloading system (FPSO), which supports Centrica’s Chestnut field, operated by Wood Group, in the United Kingdom North Sea.

Risks of Sand in Production Vessels

The presence of sand in production vessels reduces capacity and affects separation performance, which leads to deteriorating produced water quality and the carry-over of water in the oil phase. If not controlled, the sand can cause a separator to be taken offline and manually cleaned. This exposes facility crew members to process hazards and may damage downstream equipment.
Conventional sand management techniques require a shutdown of 10 to 14 days for manual sand removal, resulting in a significant delay and additional expense. Following this, oil-on-sand discharge rules may limit disposal options and compel storage of some or all of the sand.

The Online Desanding Technology

The desander is designed to allow more operator uptime and to reduce costs, compared with traditional vessel entry cleanout methods. With the new technology, there is no need to shut down production, and the elimination of manual cleaning improves safety. In addition, the technology enables operators to change their approach to planned shutdown periods. Because the system allows a high degree of cleaning to take place before shutdown, the team of technical experts needed for a shutdown can be smaller and planning for it simpler.

The system comprises a number of purpose-designed, integrated components including solids extraction, solids separation, and solids conditioning packages. A high-flow jetting unit with a recovered-liquid package recycles the separator water. The system complies with European Union hazardous area safety requirements and can be used in Zone 2 hazardous offshore environments.

The technology deploys jetting heads into vessels accurately and safely, while collecting the recovered material into process facilities in an internationally standardized container that includes a wash tank, a high-flow centrifugal pump, and an inbuilt access. All of the components operate from a single 63A electrical socket. The produced solids are washed, checked, and quantified before overboard discharge. With the assistance of the Macaulay Institute in Aberdeen, a system of field analysis kits was developed to verify overboard discharges for contamination levels in real time.

With the design’s modular concept, the existing wash tanks could be used with a separation package equipped to precondition the slurry stream. A coarse “spin filter” on the skid inlet handles sand bursts from the point of startup, significantly reducing the risk of blockage. The sand jetting nozzle is deployed by means of a solids extraction ram with rotational drive to facilitate sand recovery from the production separators.

A system of controls and process interlocks ensures that equipment operation poses minimal risk to workers and the installation. Verification studies, including a full hazardous operations analysis, were conducted in the development stage, with the support of independent engineering authorities and client process engineers.

The system’s mobility enables sand production problems to be handled cost effectively because there is no need to design or make specialist hardware or shut down the process for installation. The package is connected through the separator drain valves, which has minimal effect on the platform infrastructure. Thermographic imaging is used to monitor sand levels before, during, and after desanding. Removal of the accumulated solids restores performance efficiency to the separator and allows increased throughput, while the extracted oil is returned to the process through the platform’s hazardous drainage system.

The recovered material is subsequently removed and treated in situ with no need for onshore transportation or storage reinjection. The portable analysis kit enables onboard, real-time testing of treated sand, followed by overboard disposal where permitted.

Operators are only able to discard untreated material overboard if the level of contamination is within limits of best environmental practice. The system guarantees that discharge is washed to below 1% oil on sand, well within the UK Department of Energy and Climate Change (DECC) and United States Environmental Protection Agency guidelines. Meeting this level is vital when applying for discharge licenses. Figs. 2a and 2b show sand samples before and after cleaning.

The Technology in Operation

The system was used in eight deployments on the Hummingbird FPSO. A well was producing sand to the FPSO’s first-stage separator, posing a risk to production rates if it could not be managed. The operator considered three options: corrective well intervention measures, reduced oil production to restrict sand intake, or online desanding at regular intervals. The third method was chosen.

Over the eight deployments, beginning in 2009, 58 tonnes of sand were removed from the separator while it remained online. There was no disruption to production or ongoing operations. Each deployment was completed without harm to people or the environment. All sand was washed to below 1% oil on sand, consistent with DECC discharge requirements. Recent discharges to the sea have averaged 0.07% (700 ppm) oil on sand, which is below the 1% target figure. Removal of the sand also allowed some separator instrumentation problems to be resolved. Further deployments are planned this year as production optimization takes place.


The AOD provides a cost-effective, environmentally compliant means of removing sand from production streams in small field operations. While many separator vessels have inbuilt desanding systems, they are prone to blocking, expensive to retrofit, and require the vessel to be taken offline in the event of failure. The AOD can be easily deployed by connecting to existing drains, removing accumulated material, extracting oil from the sand, and allowing the recovered oil to be returned to the process through the platform’s hazardous drainage system. Most significantly, the technology reduces the crew size for shutdown periods, thus simplifying the planning process.

Multidomain Data and Modeling Unlock Unconventional Reservoir Challenges

Completion strategy and hydraulic fracture stimulation are important keys to economic success in low-­permeability and unconventional reservoirs such as tight sand and shale. Therefore, engineering workflows in unconventional reservoirs need to focus on completion and stimulation optimization, just as they do well placement and spacing.

The primary obstacles associated with optimizing completions in these reservoirs have been the absence of hydraulic fracture models that properly simulate the complex fracture propagation common in many reservoirs, the lack of efficient methods to create discrete reservoir simulation grids to rigorously model hydrocarbon production from complex hydraulic fractures, the lack of automated fracture treatment staging algorithms, and the lack of ability to efficiently integrate microseismic mapping measurements with geological and geophysical data.

This article details a novel approach for enabling efficient multistage completions, new complex fracture models, unstructured ­gridding-based reservoir simulation, and a comprehensive integrated workflow developed within single ­reservoir-centric stimulation design software. Though primarily focused on unconventional reservoirs, this efficient full-cycle seismic-to-­simulation workflow is applicable also for conventional reservoirs.

Seismic-to-Simulation Through Stimulation

In recent years, an increasing amount of hydrocarbons in North America have been produced from unconventional reservoirs, with a similar trend expected worldwide. The International Energy Agency projects the hydrocarbon reserves in unconventional reservoirs to be as high as 8 trillion BOE. Most of these low-permeability reservoirs must be hydraulically fractured to produce. In unconventional reservoirs, stimulation is the key step in the seismic-to-simulation workflow.

For unconventional reservoirs, detailed reservoir characterization plays a fundamental role in completion design, affecting the staging strategy, location of perforations or fracture ports, and stimulation design. Properly modeling the interaction of induced fractures with the rock fabric and coupling the hydraulic fracture geometry and conductivity with subsequent well performance is an essential element in this comprehensive workflow. The unconventional reservoir simulations include a detailed geologic description. The modeling is focused on the well and the specifics of the completion (i.e., the hydraulic fracture treatments). In contrast, this information can be significantly scaled up for conventional reservoirs, shifting the focus to large-scale reservoir behavior (i.e., multiwell and full-field simulation models).

Thus far, the oil and gas industry has lacked software that can enable such ‑seismic-through-stimulation-to-­simulation workflows. The recent introduction of such software has enabled the integration of specialized completion algorithms with fit-for-purpose hydraulic fracture models and related workflows, using detailed reservoir characterization data critical for unconventional reservoirs (Fig. 1). The completion workflow leverages a platform that enables cross validation with an Earth model developed with seismic-derived structural information; petrophysical information from logs, cores, and rock cuttings; and geomechanical data.

The capability of building multi­domain and multisource cross-­validated Earth models is leveraged to account for heterogeneity that can affect completion and stimulation significantly in unconventional reservoirs. The completion and stimulation models can be calibrated using microseismic measurements in the context of local geology and structure. This new completion solution is constructed to generate unstructured grids efficiently to represent the complex fractures in reservoir simulation models. The following sections highlight this software, which has enabled running a comprehensive single-well-optimization workflow in hours.

Completion Advisor

The vast majority of unconventional reservoirs in North America are developed using horizontal wells with multiple hydraulic fracture treatments. This approach maximizes reservoir contact and minimizes the surface footprint. These completions typically consist of 10 to 20 propped fracture treatment stages, with each stage containing two to eight perforation clusters designed to promote multiple fracture initiation points. Until recently, questions such as how many fracture treatment stages and perforation clusters are optimal, what is the ideal spacing between perforation clusters, and where is the best location for each fracture treatment stage have challenged completion engineers working in unconventional reservoirs. Optimizing the number and location of fracture treatment stages has been primarily a manual, time-intensive process, resulting in a cookie-cutter approach that does not properly account for vertical and lateral heterogeneity. Because the industry practice of geometric staging and perforation cluster placement so far has ignored the variability of rock properties along the wellbore, results from a study including production logs from 100 horizontal wells showed an enormous discrepancy in production between the perforation clusters that is likely because of rock heterogeneity. Fig. 2 presents examples from four wells that are representative of the observed phenomenon.

One important component of well performance is identifying zones of best reservoir quality and completion quality (Fig. 3). To design perforation-cluster placements in each stage, reservoir quality and completion quality are evaluated to find the best locations along the lateral. Reservoir quality implies the rock’s ability to contribute to the production, while completion quality is a measure of rocks tendency to yield under increased pressure during hydraulic fracturing. The rock parameters that define each of these qualities can be selected in the new Completion Advisor, a software with the ability to assign proportionate weighting to each parameter.

The staging algorithm for horizontal wells first evaluates the variability of reservoir properties and completion properties separately. Properties such as water saturation, intrinsic permeability, and total organic carbon will be considered as reservoir quality, while mineralogical properties such as clay and silica content, in-situ stress, Poisson’s ratio, and Young’s modulus (E) will be considered as the completion quality. The variability of these properties is separately considered and discriminated for desired values using specific cutoff criteria. The reservoir quality and completion quality derived from this exercise is assigned a binary designation of good (G) or bad (B) quality blocks, where “bad” is a relative term for rock that may still produce but is of poor quality. A composite score is derived by combining both of these qualities that can have one of the following four values: GG, GB, BG, or BB. Note that the wellbore is automatically segmented depending on the variability range of the parameters that constitute the two qualities. Different and more relevant cutoffs are applied within each of the segments.

This ranking of rocks along the wellbore provides a convenient way to automatically assign numbers of and locations of perforations as well as group similar rocks together in individual stages. The Completion Advisor, however, also acknowledges practical concerns, such as perforation spacing, to account for stress shadow effect, limited entry to ensure even distribution of pumping fluids, and start/end perforation-­cluster spacing between stages.

During the automated completion design, the software also takes into consideration the structural constraints, such as proximity of natural fractures to perforations or faults, and operational constraints, such as stage length, based on achievable rate with available hydraulic horsepower. With this novel solution, the entire process is executed in minutes or hours instead of weeks or months. The consistent step-by-step engineering approach and efficiency lends itself to multiple-­scenario analysis and helps in making decisions that are no longer stochastic. A similar but separate advisor is also available for vertical wells in differentially depleted lenticular tight sand formations. This approach provides a rigorous and repeatable solution for optimizing staging and perforation design while also rationalizing the value of the log measurements.

Complex Fracture Modeling and Calibration

Once the completion strategy (i.e., determination of fracture treatment staging, location of perforation clusters, and number of perforations in each cluster) is finalized, fracture treatments must be designed for each stage. Understanding hydraulic fracture complexity can be a critical component in the economic development of many unconventional resources. One important component in the quest to understand hydraulic fracture complexity that has been missing is the ability to model fracture propagation in complex geological environments in which the interaction of the hydraulic fracture with natural fractures or fissures is likely to result in complexity.

Models of complex fracture networks require a more complete description of the stress field than that required by simple models of vertical planar fractures. The other primary factors controlling hydraulic fracture complexity are the distribution and properties of natural fractures and the stress regime. Natural fracture description is typically derived from borehole image logs, seismic interpretations, outcrops, or cores and is represented as a discrete fracture network with seismic data-based propagation. The discrete fracture network is also correlated with curvature, azimuthal variation for orientation, and density.

The complex fracture network is strongly influenced by the interaction between the hydraulic fracture and the pre-existing natural fractures. The Unconventional Fracture Model (UFM), within the reservoir-centric stimulation design software is the industry’s first solution that explains the complex fracture patterns created with induced and natural fracture interaction and has sufficiently efficient run time for daily use. In the heart of the UFM is a criterion developed to determine ­whether a fracture crosses a frictional interface (pre-existing fracture) at different angles and principle horizontal stress differences. This criterion has been validated using laboratory experiments with excellent agreement.

In unconventional reservoirs such as shale, hydraulic fracture growth can be very complex. The UFM provides the unique ability to adequately capture the interaction between the hydraulic fracture and natural fractures. While the new software solution provides the much needed 3D visualization of hydraulic fracture in full reservoir context, it can also generate maps to show the footprint of different proppants pumped during the treatment, plots to capture total fractured surface area created over time, and the effect of stress shadow in both the intrastage and interstage. The UFM, however, needs to be validated in these complex environments using microseismic measurement data (Fig. 4). The single canvas concept is leveraged to overlay processed microseismic data on ­model-generated fracture geometry and the workflow to perform model calibration.

Reservoir Simulation

Optimizing well spacing, well placement, and fracture treatment design efficiently requires seamless integration with production modeling. Thus far, representing complex fractures accurately and effectively to properly evaluate the effect of hydraulic fractures in unconventional reservoirs has been a challenge. Hydraulic fracture treatment design requires reliable production forecasts to evaluate the effect of treatment parameters on well performance (e.g., proppant type, size, and amount; fluid type and volume; and injection rate). Discrete gridding of the hydraulic fracture in a numerical reservoir simulation model is the most robust and flexible approach but can be time-consuming and cumbersome if the process is not automated.

Automated gridding of multiple planar hydraulic fractures in horizontal wells has been possible for many years, but the application of these integrated approaches has been limited. Coupling complex hydraulic fracture models with reservoir simulation using algorithms that automatically develop discrete reservoir simulation grids to rigorously model the hydrocarbon production from complex hydraulic fractures completes the seismic-to-simulation workflow for unconventional reservoirs.

As with hydraulic fracture models, the reservoir simulation models need to be calibrated using actual production data to ensure the production forecasts are reliable. This calibration process is intimately linked to understanding hydraulic fracture performance—specifically fracture conductivity, effective fracture surface area, and stimulated ­volume (Fig. 5).


In this article, we have discussed the deployment of a unique, reservoir-­centric stimulation design software that offers several previously unavailable components. The Completion Advisor enables the connection between data and information by linking measurements to actual stage and perforation design decisions through an efficient yet rigorous engineering approach. The complex hydraulic fracture simulation model captures the physics of hydraulic/natural fracture interactions and is efficient enough for day-to-day fracture design and production evaluation. The automated grid generation for complex hydraulic fracture geometries is seamlessly coupled with a reservoir simulator, allowing efficient fracture treatment optimization and field planning.

Finally, the software solution enables the seismic-to-stimulation-to-­simulation workflow that is necessary for single-well optimization in unconventional environments in a single platform. This eliminates the need to move data from one application to another, address data formatting issues, learn multiple and various software tools, and address problems at the interface that can easily become a bottleneck for asset teams operating in unconventional ­environments.