Comprehensive Workflow for Wireline Fluid Sampling in an Unconsolidated Formations Utilizing New Large Volume Sampling Equipment

WHOC11-343 Comprehensive Workflow for Wireline Fluid Sampling in an Unconsolidated Formations Utilizing New Large Volume Sampling Equipment S. KVINNSLAND, M.BRUN TOTAL E&P Norge V.ACHOUROV, A.GISOLF Schlumberger This paper has been selected for presentation and/or publication in the proceedings for the 2011 World Heavy Oil Congress [WHOC11]. The authors of this material have been cleared by all interested companies/employers/clients to authorize dmg events (Canada) Ltd., the congress producer, to make this material available to the attendees of WHOC11 and other relevant industry personnel. Abstract Wireline formation testers (WFT) are well established in the industry, with many applications ranging from pressure profiling and PVT sampling to interval pressure transient testing and stress testing. This paper outlines how the latest WFT technologies are used to address key sampling limitations: acquiring PVT and especially large volume samples in the wells drilled with oil based mud (OBM) in unconsolidated formations. Such samples are very critical for flow assurance studies and production facilities design. While sampling from unconsolidated reservoirs, small solid sand particles can be produced from the formation. Depending on the pressure drawdown and level of the rock weakness, such sand production can lead to tool plugging before representative samples are taken. Cleaning action is then required which is time-consuming and results in interrupted flow periods. This makes filtrate clean up very difficult and it is particularly disruptive for large volume sampling. Technology developments used to address this problem include slow rate pumps, focused and large area probes and filters. New large volume sample chambers have also been developed. Successful utilization of these technologies, particularly in unconsolidated formation requires a comprehensive workflow. This paper describes such a workflow including an example of sampling in highly unconsolidated sand, saturated with biodegraded fluid. Several PVT sampling stations were required. Conventional welltesting was not planned, so about 70 liters of low contaminated oil were needed for various fluid studies. To obtain such a sample in an OBM well a long and uninterrupted flow period is needed. The first WFT configuration attempted included an ultra slow pump for PVT sampling. One of the acquired PVT samples was used for the on-site filtrate contamination analysis. This analysis helped to design the next run, which resulted in the first successful large volume sampling in OBM environment with the new WFT chambers. Introduction Precise and accurate knowledge of fluid properties (see Table 1), were essential for the decision on commingling production from the Hild filed in the North Sea. Reservoir description and Formation Testing and Sampling Objectives. 1

The discovery well of the Hild oil accumulation, 30/7-2, was drilled in 1975. The well encountered a gas cap overlaying a 20 meter oil column. The Frigg sands formation is relatively well known, both from the 30/7-2 well data and also from neighbor fields data (Frigg field and Nuggets). The reservoir is known to be high porous and high permeable unconsolidated sand. A DST was carried out in the oil zone and an oil sample was taken from the test separator for PVT analysis objectives in the well 30/7-2. The PVT analysis from 30/7-2 showed heavily bio-degraded 5 cp viscous oil with a API of 22 degree. No further fluid studies were done at the time (and no oil samples remains). Therefore uncertainties on the fluid characteristics remained. A Hild field development will include Frigg oil and Brent oil and condensate in the development design. With reference to the PVT analysis from the 30/7-2 well the nature of the Frigg oil characteristics and the degree of biodegradation were worrying with regards to the process design, flow assurance and export solutions. Therefore the TOTAL architects engineering group requested Geosciences group to provide oil from the Frigg oil accumulation for production chemistry studies. The critical issues for design were among others; crude assay and TAN value for oil & project valorization, viscosity and emulsion characteristics for process facilities and export routes, host and terminal screening etc. Further, the oil characterization was a key missing data for a potential combined Frigg oil and Brent gas development design where fluid compatibilities testing of the fluids from the two Hild field reservoirs (Brent and Frigg) were necessary to allow for commingled production. A general strategy was carved out for the Hild appraisal well. Main objective was to derisk the geosciences uncertainties on the deeper Brent reservoir where gas and condensate volumes representing possibly 80 % of a Hild field development. But the Frigg oil accumulation was decided as a secondary objective for the appraisal well and an attempt to meet the APP request with regards to oil sampling for production chemistry issues should be done. The well trajectory should, if feasible, be designed to allow traversing the ileocene Frigg sand formation within the oil pool without compromising the Brent well trajectory below. Further, in the context of Frigg oil being a secondary target in the overall Hild field strategy, the well design was to allow for wireline formation testing and sampling only in the Frigg reservoir. If the wireline formation testing and sampling failed in terms of the required sampling volumes in the Frigg reservoir due to the lack of sufficient oil column and/or sampling conditions (emulsions, mud contamination etc.) a dedicated well on Hild Frigg reservoir with a DST should subsequently be considered. Frigg fluids sampling requirements are shown in the Figure 1. Basic fluid analysis required three 450 cc PVT samples from three stations across the oil column, one station in the gas cap and one station in water zone in the Frigg reservoir to be taken with wireline formation tester. For the production chemistry analysis the requirement for sampling volumes was significantly higher, with 5 liters requested for the general production chemistry analysis, 10 liters for crude assay studies and 40 liters of oil from the Frigg reservoir for the specific dynamic separation test. All these volumes had to be taken from the middle of the oil column with the lowest possible contamination level. Normally such a samples are acquired from the surface separator during the DST testing and/or production testing. Historically such large volume sampling was not possible with wireline deployed sample chambers. Although there were large volume sample chambers available with wireline formation testers (with the volume about 22.7 liters) they were not designed to be placed in the flowline of wireline formation tester tool for low contamination sampling with use of downhole pump. However, large volume sample chambers have recently been developed that can be placed anywhere in the flowline stream of a wireline formation tester. They can be used with the standard probes and focused sampling probes. This allows the capture hydrocarbon samples that do not suffer from property changes brought on by mud filtrate contamination. Sampling from highly unconsolidated sand, saturated with heavy bi-degraded oil, requires detailed planning. This includes researching lessons learned from the past and exploring new technology solutions available. Field experience Field experience has shown that when sampling fluids from such formations, sand grains tend to become mobilized and flow with the fluids being sampled 1,2. When significant amount of solids and grain sands are mobilized it could result in plugging the formation tester probes, flowlines and downhole pumps. Multiple trips to surface might be required to clean or even replace equipment or equipment parts. There are two approaches to successful sampling in unconsolidated formation utilized in this paper; Prevention and mitigation. The first approach is to prevent solids mobilization by reducing the pressure drawdown that the formation is exposed to. The second approach, to mitigate the effects of mobilized solids, involves the use of various filter systems. This approach allows pumping of formation fluids long enough for mud filtrate levels to be 2

reduced to acceptable levels before plugging the formation tester filters. Technology/Solution Variety of Wireline Formation Tester modules used for sampling in unconsolidated sands is discussed in details in reference 1. As stated, prevention of solids mobilization could be achieved by reducing the pressure draw down the formation is exposed to. This can be achieved by increasing the inflow area of the probes. For any fixed rate flow through a permeable media, increasing the area open to flow will effectively decrease the differential pressure across that media. There are a large amount of probe sizes available in the industry. Many are detailed in the references 1-4. The range spans, but is not limited to, standard, large diameter, extra large diameter, focused, large area and elliptical probes. Drawdown can also be controlled by minimizing and controlling the flow rate generated by the pump. A formation tester pump module consists of an electric motor, a hydraulic pump and a displacement unit. To reduce the pump rate ranges, both the hydraulic pump and the displacement unit can be changed. References 1 and 2 details such available pump equipment, including standard, high pressure, extra high pressure and extra extra high pressure displacement units of variable displacement hydraulic pumps, fixed displacement and twostage hydraulic pumps. Correct selection of this equipment allows controlling the pump rate over the wide range. Mitigating measures against migration of solids consists of the use different probe filters and inline filter modules. Most of the previously mentioned probe barrels can be fitted with different sizes filters or even a gravel pack filter. In addition to this the Martineau probe contains a filter chamber behind the probe. Alternatively the filter can be removed from the probe altogether and an in-line filter can be placed between the probe and the pump. Again references 1 and 2 contains details of various options available. Sampling can be a time consuming operation at the best of times, but never more so then when large volumes of formation fluid are required from an unconsolidated sand environment in a well drilled with oil based mud. This is the case presented in this paper. Therefore managing the plugging risks is even more critical than normal. Clearly with such a wide range of equipment available for sanding prevention and mitigation measures, there is a need for a comprehensive workflow detailing what equipment to use for this environment and in what order. Comprehensive Workflow for Wireline Formation Testing and Fluid Sampling applied in the Hild well. In October 2009 an appraisal well in the Hild oilfield was surveyed in the Frigg reservoir using a wireline formation tester for pressure and mobility profiling, PVT sampling of gas and oil as well as for water sampling and large volume oil sampling. The unconsolidated Frigg reservoir is saturated with biodegradated oil with a gas cap on the top of the reservoir and water leg below the oil column. Full and comprehensive risk assessment, job design, and several contingency run options were made before the job and this led to successful operation. This required utilization of a large range of wireline formation testing and sampling technologies, along with the use of new large volume sample chambers to meet the sampling volume requirements. A comprehensive formation testing and sampling strategy workflow is shown on the figure 2. Wireline runs are enumerated on this plot in the following order. Run 1 (it is not shown on the workflow figure) contained petrophysical logs. Run 2 was for formation pressure and mobility profiling. Run 3 was for contingency pretesting and mainly designed for gas and oil PVT sampling as well as for water sampling. Run 4 was dedicated for the large volume oil sampling. The workflow generally consisted of the three runs with one or more contingencies for each run. The various contingency runs workflow can be read from figure 1. We will give details of the reasoning behind the workflow used. Pressure and mobility information will give an estimation of the fluid contacts and optimize the sampling depths. To minimize the long pumping time, sampling depths are ideally chosen close to the flow barriers, so the amount of filtrate flow could be minimized due to the decreased vertical permeability. The mobilities obtained will help to determine where the best mobility can be found and help balance this to the proximity of shoulder beds. Pressure testing Run 2 should also give an initial indication of the probe plugging. The pressure testing and mobility profiling run was planned with two standard probe modules one conventional and one large diameter probe. The smaller standard probe might seal better, but the larger area of the large diameter probe will have less chance of plugging. If the probes in this run all plug, there is a contingent Run 3, combined with PVT and water sampling that will be used to aquire the remaining pretests. These contingent probes included an extra large probe to reduce draw downs and a Martineau probe to filter formations fines. Run 3 was planned for hydrocarbons PVT and water sampling. Base configuration for this run also has two standard probe modules, one conventional and one large diameter; two variable displacement pumps one with standard displacement unit and one with the high pressure displacement unit; and fluid analyzers for fluid profiling and contamination monitoring. This first sampling run has several 3

contingency options in case probes or downhole pumps are plugged. First contingency option for the PVT sampling run is based on changing the probe types from standard probe and large diameter probe to extra-large diameter probe. This is to increase the inflow area and thus to minimize the pressure drawdown. If these prevention measures fail there will be a Martineau probe available with a filter chamber behind the probe and lastly, an inline filter module Inline filters have a large filter volume and can filter out large volumes of solids. Pumps selection for these two contingency runs is based on decreasing the pumping rate range by increasing the variable displacement pump displacement unit volume or using the fixed displacement pump. One more prevention measure was implemented which was a special sample chamber configured as an exit port. Closing this exit port while moving the tool will help to prevent solids, which may suspend in the mud, from entering the flowline and plugging the pumps. Also the bottom pump was used for the initial pumping. In case if near wellbore damage could cause the high drawdown and thus potential plugging of the pump at the initial stage of pumping use of the pump placed on the bottom side of the tool below the probe and pumping down could help to save the main sampling part from the plugging before the main part of cleaning process is achieved. These probe, filter and pump contingencies are spread out over two contingent runs, 3.2 & 3.3, Figure 1. They all provide sanding prevention measures by minimizing the pressure drawdown while pumping and thus potential sand grain production. In addition there are mitigation measures through the use of filters. Next, run 4, is large volume sampling with three six gallon sample chambers. It also has two contingency runs. One of them is similar to the contingency Run 3 for PVT in terms of probes, pumps and filters selections. An additional contingency includes the focused sampling probe 4 with extra high pressure variable displacement pump on the guard side and ultra slow rate fixed displacement pump on the sampling side. Use of Quicksilver probe was planned in case the single probe sampling attempts would require unacceptably long mud filtrate cleaning times. Ultra low contamination was critical for the large volume samples. Focused flow, coupled with ultra slow pumps for low draw down, is expected to result in lower contamination samples than samples obtained with a single probe flowing at the same rate. Case Study Results Figure 3 shows the depth view of the tested interval with the formation pressure (first track), fluid fractions from the downhole fluid analysis (track 2), mud pressures (track 3) and drawdown mobilities (track 5) along with the open-hole logs shown for reference. Formation fluid gradients of gas, oil and water have been drawn over a number of pressure tests to estimate the fluids contact depths. It is interesting to note that formation pressure gradient calculated from the points acquired in the oil zone may suggest an increasing trend in in-situ fluid density. This could be explained by the bi-degraded nature of formation oil. Selecting sampling intervals for the PVT run was based on formation fluid mobility and for and the large volume sampling run it was mainly based on the suggested proximity to the flow barriers around the probe (to minimize the long expected pumping time) During the sampling run 3.1 standard 450 cc PVT sample bottles were acquired from the three depths in the oil zone along with overpressured single phase 250 cc sample bottles from the gas zone and 1 gallon sample chamber from the water zone. After this sampling run, as soon as the sample bottles arrived at surface, one of the PVT bottles, which has been taken from the middle of the zone surrounded by the impermeable streaks, was used for the on-site filtrate contamination analysis. The resulting sample contamination level was less than 4%. Which agreed quite well with downhole fluid analysis contamination modeling and considered to be a good sample, which allowed the operator to select run 4.1 for the large volume sampling. This run included three six gallon sample chambers as per the described workflow shown on the figure 2. The pressure drawdown (difference between formation and flowing pressure) on this station was less than 0.4 bars which has been achieved with the flow rate varying about 0.9 cc/sec from the zone with drawdown mobility of about 670 md/cp. Figure 4 shows the change of GOR and strain gauge temperature while pumping on the large volume sampling station of the run 4.1. Use of fixed displacement pump for main pumping and sampling during this run allowed to achieve the slowest rates of 0.66 cc/sec with the drawdown within 0.3 bars in the zone with mobility of about 1200 md/cp. As it shown on the figure 3, GOR was slightly increasing during the pumping. Figure 5 shows contamination prediction modeling where both color and methane models also show an average mud filtrate contamination on the levels of about 5% and more at the end of the pumping. However based on the contamination results obtained from the previous run and contamination modeling based on the change of fluid color it has been decided to fill all three 6 gallon sample chambers at about 3.3, 7.4 and 11.7 hours respectively. Total station time was about 15.7 hours with the filling time for each of the six gallon sample chamber of about 3.6 hours. Later it has been measured in the lab that OBM contamination of the samples acquired during this run is less than 1%. Conclusion 4

The requirement to get quantities from the sampled oil for various studies has induced important priorities rules on the preciously sampled volumes. Use of comprehensive workflow to acquire large volume samples with wireline formation tester in un-consolidated sand helped to acquire about seventy liters of formation oil. This was achieved in a one tool descent in hole with the samples contamination level below 1%. The sampling operation has been proven successful, and the precise characterization of the oil allowed for significant design updates the Hild project. From the crude valorization, production chemistry and flow assurance point of view the oil show characteristics that became a game changer for the field development. A high TAN value of the oil has given constraints on export options. The screening process to optimize the design for oil treatment and export alternatives (surface and host compatibility studies etc.) have imposed significant analysis needs and confirming the need and use of the totality of the sampled volumes. Acknowledgement The authors would like to thank TOTAL and Schlumberger for the permission to publish this paper. References 1. Jackson, R.R, De Santo, I., Weinheber, P., Guaragnini, E., Specialized Techniques for Formation Testing and Fluid Sampling in Unconsolidated Formations in Deepwater Reservoirs ; SPE paper 120443 presented at the 2009 SPE Middle East Oil and Gas Show and Conference held in the Bahrain International Exhibition Center, Kingdom of Bahrain, 15-18 March 2009 2. Jackson, R.R., Santo, I.De, Weinheber, P., Guaragnini, E., Innovative approach to Formation Testing and Fluids Sampling in Unconsolidated Formations: A success Case From Offshore Africa, Nape paper prepared for presentation at the 2008 Nigerian Association of Petroleum Explorationist s Conference and Exebitions held in Abuja, Nigeria, 16-20 November 2008. 3. Wireline Formation Testing and Sampling, Schlumberger Educational Services, SMP-7058, Houston, USA, 1996 4. Weinheber, P., Gisolf, A., Jackson, R.R., De Santo, I.: Optimising Hardware Options for Maximum Flexibility and Improved Success in Wireline Formation Testing, Sampling and Downhole Fluid Analysis Operations, SPE paper 119713 presented at the SPE 2008 Nigeria Annual International Conference and Exhibition held in Abuja, Nigeria, 4-6 August 2008. 5

Analysis Criticality Why Min. OBM contamin. Total Acid Number High Export/host impact Not an issue Critical water cut and viscosity Moderate Bio degradation Low Separator sizing + flow assurance Mobility/reserves ==> Project economy Not an issue Crude Assay/valorization High Project economy 15-20 % Impact if not achieved Penalties from host or worst case: Impossible to blend Over-design of sep. ==> Cost & space Not an issue Project economy Penalties on sale/host impact/host blend acceptance Sample quantity (ltrs) 2.5-4 15-20 Wax Moderate Asphalthenes Low Flow assurance - pipeline Process design, Flow assurance - pipeline 10-15 % 10-15 % Pipeline design, Pigging freq- & chem. Inj. Skids Process & pipeline design Emulsion properties Moderate Sizing of separator + flow assurance < 5% Operability, Over-design of sep. 40 Compatibility with Brent cond. (and 3rd party host liquids ) High Flow assurance - pipeline < 5% Concept & process design The samples marked by are done with the same sample Table 1 Hild oil (Frigg) fluid properties study requirements. Figure 1 Fluid sampling strategy. 6

Figure 2 - Comprehensive workflow for sampling in unconsolidated formations. Tool Mnemonics. PS Conventional probe, PS(LD) Large diameter probe, SC Sample chamber configured as an exit port to prevent formation tester flowline from plugging and to allow reversing the downhole pumps for downhole cleaning, PO (STD) Standard displacement unit pump HY Probe hydraulic unit LFA Fluid analyzer PO (HPDU) High pressure displacement unit pump CFA Compositional fluid analyzer SC1G One gallon sample chamber MS Multisample carrier MPSR PVT sample bottle SPMC Single phase sample bottle SC6G Six gallon sample chamber PS (Martineau) probe with large area filter, which could be also run with gravel-pack PO (FDU) - Fixed displacement volume pump PO (XHPDU) - Extra high pressure displacement unit pump PQ Quicksilver probe SC6G Six gallon sample chamber MS (Sand Trap) Inline filter module 7

Figure 4 GOR and Temperature change while pumping on the large volume sampling station Figure 3 - Formation pressure and mobility profile. Figure 5 Methane and color contamination prediction modeling 8