Abstract

The Azeri-Chirag-Gunashli (ACG) is a giant field located in the Caspian Sea, Azerbaijan. The field is operated by BP on behalf of Azerbaijan International Operating Company (AIOC). Wells are drilled and completed penetrating multilayers of weakly consolidated sand with low unconfined compressive strength. The typical sand-face completion is Open Hole Gravel Pack (OHGP), Stand Alone Screen (SAS), Expandable Sand Screen (ESS) and Cased and Perforated (C&P). After a period of time, when a well does loose it's sand control integrity, the wellbore has the potential to fill up with sand and create a production deferment. A viable option to revive these wells is by cleaning the sand out of the tubing to gain access and perforate the upper-zone.

This paper describes the success story of a challenging sand clean out and return to production in one of the wells in the ACG field. The job utilized an integrated clean out system from downhole to surface. A hydraulically actuated, switchable circulation sub was used with a real time downhole measurement system. While running into the well, forward jets of the sub helped penetrate and suspend sand and while pulling out of the well, the sub was switched to a low resistance, backward jetting mode which sweeps sand more efficiently. Adding real-time downhole measurement to the system enhanced the performance of the sub ensuring it was functioning correctly as per plan. A sand monitor was used in the surface return line to continuously monitor the returned sand. Design software was used to support the operation from start to finish and real time data comparisons to the software predictions created additional assurance of safe operational steps. Following the successful cleanout, the plan was to isolate the lower sand producing zone and perforate the well across a higher oil bearing zone. A live well deployment system on electrical coiled tubing was selected to perforate over 300m of the pay zone in a single run. This system was selected due to its capabilities of deploying and reverse deploying the extra-long perforation guns with positive wellhead pressure and completing the job in one run. Combination of the electrical and mechanical features of the coiled tubing added additional value to the operation as it made it possible to correlate depth based on CCL and gamma ray, displace kill weight fluid to base oil to create underbalance conditions and monitor the well pre and post perforating with CT in the wellbore.

The intervention was successful and the well was put back into production. This case serves as a reference for future sand clean outs and CT conveyed live well perforation jobs in the region.

Introduction and Job Overview

Well C on the C platform located at Azeri-Chirag Guneshli field is an oil producer completed in May 2004 with 4-1/2" expandable sand screens across the production zone. In 2010 a large amount of sand invaded the wellbore and started to be produced to surface. The well was choked back and able to produce sand-free at a reduced rate for a while before it died completely. The well remained shut-in until a cleanout could be performed in 2013. It was decided to remove the sand fill or bridge out of the wellbore, isolate the lower sand producing zone and perforate a different production zone higher in the completion.

Well C can be characterized as being drilled near horizontal (fig.3) with a large completion size (fig.4) and long tangential section around 600m. In order to revive the well several engineering approaches were considered to successfully and safely accomplish this goal.

The final decision was taken to utilize the CT approach and complete the operation in three stages: removing sand fill or bridge, isolating lower production zone and perforating the new zone. Electrical CT was selected as the best candidate to accomplish multiple tasks. A specialized sand cleanout process, electrically-activated plug setting and live well perforator systems were deployed on CT with successful results.

The CT cleanout system consisted of a specialized software module to calculate the required dynamic forces to bring sand to surface, predict hydraulic behavior of the well and optimize the wiper trip speed and hole cleaning time for the operation condition. It was also used to guide the field personnel with regard to CT operational parameters.

A selective bi-directional sub was used to penetrate through the tight-packed sand by means of forward oriented high impact jets (fig.2). Transporting fluidized hard particles to surface was accomplished by reverse oriented swirl type flow through up-hole facing nozzles (fig.1). The tool is constructed to work on hydraulic power and achieve 100% of flow going through either forward or reverse nozzle at any one time (fig.2). Real time electrical CT system made it possible to carefully monitor bottom-hole conditions and it was helpful in identifying the BHA functionality downhole. The surface sand handling system consists of sand detecting clamp-on devices as well as a sand cyclone tank to separate the produced solids from fluid before sending the fluids to the production test separator.

Figure 1

Sand Cleanout with Switchable Sub

Figure 1

Sand Cleanout with Switchable Sub

Figure 2

Switchable Sub in action

Figure 2

Switchable Sub in action

Figure 3

3D View Well C Trajectory

Figure 3

3D View Well C Trajectory

Figure 4

Well C Completion Configuration

Figure 4

Well C Completion Configuration

A plug to isolate the lower zone was set downhole with a setting tool and was activated via an electrical motor that opened a nitrogen chamber to stroke the plug setting piston. The electrical signal was sent to the setting tool via the electrical CT.

The well was then perforated using a combination of live-well deployment BOP technology and an electrical firing head, triggered using the electrical CT.

The main reasons for deciding to perforate on electrical CT were:

  1. Coiled tubing was able to handle the weight of over 300m of gun sections which allowed completing the operation in a single run versus 16 runs using an e-line tractor.

  2. The combination of mechanical and electrical capabilities made it possible to reliably transfer an electrical signal to the BHA and therefore provide accurate depth control while at the same time delivering improved safety at surface and greater flexibility with both electric and hydraulic redundant firing options.

Required Resources

Sand cleanout software module

The proprietary job modeling software drove the need for the limited length "bites" into the sand fill and also the frequent short wiper trips to re-distribute sand across a longer tangent interval. This was then to be followed by performing the full wiper trip to surface. The design had to be mindful of the low bottom-hole pressure and large amount of fill which in fluidized form would create sufficient hydrostatic head to initiate circulation losses. The limited capacity of handling solids at surface was another reason of careful approach. The software predicted a large volume of sand would hit the sand separator within a short period of time when the wiper trip approached +/-1000m MD. Another area highlighted by the software was the failure of fully cleaning the wellbore by circulation alone due to the deviation angle and large ID – confirming conventional CT cleanout approaches would not be successful. Therefore a further optimization was implementing the specialized switchable sub, yielded positive results indicating improved cleanout efficiency. The required fluid pump rate and optimum wiper tripping speed at said pump rate were also calculated by the software. One final output of the software was calculating the amount of friction reducer required to achieve optimum pumping rate.

Bidirectional switchable sub

It was identified during pre-job modeling that frequent short trips and accumulative wiper trips to surface would be the most effective way to clean the sand from inside the completion. It was also predicted that the job objective could be accomplished by creating localized high impact stress from forward jets to break any hard pack and circulate sand further up-hole by means of high pump rate reverse oriented flow. By pumping at the rates 1.4 bpm or below, forward jetting action was selected to impact the fill with high forces and made it possible to penetrate into the sand. Increasing the rate above 1.4 bpm diverted flow to reverse facing jets. Flow through reverse nozzles made it possible to increase fluid rate because of decreased pressure drop across the larger nozzles. The rate could then be increased up to 3.0 bpm while keeping coiled tubing pumping pressure within acceptable limits.

Electrical CT System

The electric CT system consists of an intelligent motor-head assembly (MHA) that combines standard MHA functionality and houses internal and external pressure sensors, temperature sensors and a CCL. Communication with the BHA is established through an insulated mono-conductor cable, which is a 1/8" OD corrosion resistant alloy tube housing an insulated electrical conductor. In summary the following can be said regarding the value gained by having electrical real time communication with the BHA.

Real time communication with the BHA facilitated a complete understanding of the bi-directional sub functioning (forward or reverse jetting). Changing downhole wellbore pressures provide a quantitative feeling of the circulated fluid being loaded up with sand and allows a judgment call to be made about the optimum length of the penetration into an otherwise blind bed of sand. The wellbore pressure at the BHA can also be used as an accurate indication of breaking through a plug or bridge and therefore prepare the crews for partial or total losses.

Depth correlation was performed based on the CCL reading and was helpful to identify if the tag depth was the same as on the previous run and therefore recognize if any sand influx from the formation had occurred. The accurate depth also helped to identify key areas of the completion where geometry changes existed and were therefore likely sand bridging locations.

Once the cleanout was completed the electrical CT system was utilized to set mechanical plugs to isolate the existing production zone with no need for e-line tractor intervention. Electrical capabilities were also utilized to activate the firing head to detonate the perforating guns downhole.

Clamp On Sand Sensor

The sensor measures sand/particles through passive ultrasonic technology and detects the ultrasonic signal that is generated by particles impacting on the inside of the pipe wall. Sensors installed on the return lines provided an indication of the sand rate at surface. This reading was compared against the software modeling results to optimize subsequent cleanout bites.

Live Well Perforator Deployment System

The Live Well Deployment System is used to deploy or reverse deploy Expendable Hollow Carrier (EHC) perforating guns from a live well without the need to kill the well. For the subject operation the system is used in conjunction with electrical CT and a specially arranged blow-out preventer (BOP) and pressure control stack. The deployment system consisted of special gun connectors not requiring rotational action to be made up or broken out, a running/retrieving BHA, electrical and hydraulic firing head assemblies, standard CT quad BOP dressed with the special deployment rams to achieve connecting / disconnecting of the gun sections under live well pressure. Gun sections are made up by a simple snap latch mechanism with no requirement for rotation making them extremely robust. Perforator sections are disconnected by functioning the deployment BOP rams to compress the latch keys and allow straight pull to release without any need to generate rotational movement. This greatly simplifies the functionality and is a significant advantage of the system.

The 300m long perforating guns were deployed in the well and were made up to the running and perforating BHA, comprised of a CT to e-line electrical adaptor, PFC - GR depth correlating device, electrical firing head and 3-3/8" perforating guns. The guns were connected using 22 deployment connectors, spaced every second gun. This system allowed the guns to be correlated on depth, displace the well to base oil and then fire the guns electrically.

Job Execution

Slickline drifted the well and top of fill was encountered at 3096m MD. Slickline bailer recovered emulsified solid/oil/water mixture consisting of 26% hard particles. Next step was to rig up CT to clean out the wellbore.

The 1st RIH with CT was to the top of the slickline holdup depth at 3096m MD and circulated the wellbore fluid to slick sea water. Multiple penetrations were made into the fill and short wiper trips commenced in an attempt to redistribute sand equally along the wellbore. Final wiper trip commenced from 3190m back to surface in reverse jetting mode. When the CT BHA was above 2000m MD, sand traces started to emerge at the sand detectors. When sand was no longer recorded by the sand detectors, pumping was stopped and the CT was prepared to RIH for the next run.

The 2nd RIH with coil was able to cleanout sand from 3190m MD to 3250m MD while keeping the same short bite and sand redistribution approach. CT pulled out of hole in reverse jetting mode from the deepest point reached and stopped at 240m based on sand detectors no longer indicating sand.

The 3rd RIH with CT was able to clean sand from 3250m MD to 3330m MD while maintaining the same careful penetration tactic and keeping with the same sand redistribution methodology as previous runs. CT was POOH in reverse jetting mode and stopped at 500m MD based on near zero sand indications at the surface sand detectors.

Run #4 and #5 were conducted in a similar manner, penetrating a further ~ 100m of fill during each run before making a wiper trip close to surface in the reverse jetting mode.

Run #6 was again similiar with regard to bite size and wiper trip philosophy, but this time the CT had to be pulled all the way to surface based on the sand detectors continuing to indicate sand while the BHA approached surface. It is worth noting that the volume of sand unloaded from the well was recovered in two distinct batches during this run. Once the initial spike of sand recorded was cleaned up there were no signs of particles circulated out for a few hours whilst still circulating and POOH on the wiper trip. Continuing to pull CT closer to surface brought the second batch of sand to surface and the surface detectors indicated sand had stopped returning in the fluid by the time the CT was pulled up to a depth of 111m MD.

Run #7 was conducted in the same manner and the total clean out depth recorded at this stage was 3900m MD. While pulling CT out of hole, slight traces of communication with the reservoir were noticed at the real time downhole pressure sensors. When CT was at depth of 3000m MD more losses were recorded downhole and finally when CT reached 2760m MD further losses occured indicating no further sand bridges remained between surface and the top of reservoir. CT was then POOH and parked at surface.

The next step was to bullhead any remaining sand in the tubing back into the reservoir to ensure the wellbore was clean to deploy a plug and perforate. CT was RIH while pumping through the CT BHA in forward jetting mode and down the 7" production tubing. It was possible to reach the planned plug setting depth during this stage with the CT.

The next operation was a dedicated drift run to ensure it would be possible to reach the planned plug setting depth with the bridge plug assembly. Plug setting depth of 4750m MD was achieved with no tag encountered.

Three separate isolation plugs were set at the planned depth to ensure isolation of the lower zones. To activate the setting assembly of each plug, an electrical signal was sent through the electrical CT to the setting tool. Clear surface indication of the plug setting downhole was recorded on the first setting attempt each time.

CT exited the well after the last plug setting run and the well was ready for the perforating stage of the operation. The pressure control stack was dismantled and replaced by the dedicated live well deployment pressure control stack. The deployment stack-up drawing is illustrated in (fig 28). The lower stack mainly consisted of a deployment BOP with a set of locate (no-go) rams for locating the gun connector, slip rams to hold the weight of the guns when disconnected from the CT and disconnect rams for disconnecting two gun connectors under pressure. Manual and hydraulic gate valves with a flow-cross in between were also rigged up to monitor the hydraulically activated gate valve with an annular BOP rigged up on top with the main purpose to centralize the BHA if encountered difficulties with retrieving the gun connectors.

Since it was the first time live well deployment had been carried out on the platform a dummy gun deployment trial commenced prior to making up the live perforating guns. CT was rigged up and made up to the dummy guns already deployed into the pressure control stack. The dummy guns were then run into the deployment BOP and function tested to ensure proper latching and unlatching could be achieved. Once complete, the dummy BHA was rigged down and started to deploy the actual guns into the well. The tested plugs in the well and monitoring of the kill weight fluid in the wellbore were the two well control barriers at this stage and allowed the guns to be dead well deployed without using the deployment BOP system. The full string of guns were hung off at the deployment BOP and the firing assembly was made up on the CT consisting of the electrical CT BHA, wireline cross over, PFC-GR tool and firing head. The CT injector was then made up to the stack and integrity of the system confirmed.

CT started to run in hole with the perforating BHA from 312m (tip of the Gun) taking displacement to the platform trip tank to be able to monitor displacement. While RIH, weight checks were performed every 300m as CT operating limits were being stretched given the weight of guns and the wellbore geometry. Once the perforating interval was reached, correlation passes were performed to confirm correct depth. With CT at depth, the well was displaced to base oil to create an underbalance condition before firing the guns. A final correlation pass was made to confirm GR correlation before the guns were fired. Successful surface indication (fig 26) of the gun firing was noted and CT started to pull the spent guns to surface without any signs of overpull. Once at surface the CT was spaced out to place the first connector across the deployment BOP. The disconnect ram was activated and the firing head BHA pulled free above the gate valves. The stack gate valves were closed in and the 1300 psi WHP was bled off in between the hydraulic and manual gate valves and from above the manual valve. The firing head toolstring was retrieved and changed out to the run-retrieval BHA. Standard reverse deployment procedure was adhered to in the following sequence to complete retrieval of all remaining gun sections without incident.

  • Makeup the run-retrieval BHA and stab into lubricator and pressure test the stack

  • Equalize the pressure above double gate valve barrier and open the valves

  • Run in hole and latch into gun section hanging on deployment BOP

  • Perform pull test to make sure male connector firmly latched into female connector

  • Back off the slip rams and locate rams to release the BHA

  • Pick up the BHA to the next connector section and activate locate ram on connector locate neck and tag it to make sure connector placed correctly across deployment BOP

  • Slack off weight on locate ram and activate slip rams, perform pull test to confirm slip rams holding BHA firmly

  • Activate disconnect rams and pick up to release male part of the connector

  • Pickup above gate valves and isolate wellbore, bleed pressure off the stack and disconnect stack at the quick connect sub

  • Retrieve spent gun sections and repeat above procedure until entire gun string is out of hole

Job Evaluation

Having a specialized cleanout software package, switchable cleanout tool between 100% forward and 100% reverse jetting and real time downhole measuring system at the BHA resulted in a successful job without any major incidents and with all the objectives set out in the scope of work being achieved.

The following can be summarized regarding the job;

  • The sand fill/bridge was removed and full bore completion access re-gained to the required depth.

  • The integrated approach of a switchable jetting tool along with real time BHA readout data helped in optimizing the overall cleanout process and gaining a full understanding of what was happening at the BHA depth at any point in time.

  • Lower zone isolation plugs were set and tested at the designated depth.

  • Time and cost effective live well perforating system utilized with no recordable downtime. Implementation of live well deployment system saved excessive runs with an e-line tractor, or the atlternative of killing the well using costly and damaging fluids..

The risk was understood prior to starting the cleanout that breaking through a sand bridge in the well bore and establishing communication with the reservoir could cause a total loss of fluid returns due to low reservoir pressure. Reservoir pressure was 1300 psi less than hydrostatic pressure of the fluid column. Upon reaching the slickline holdup depth, pressure sensors at the BHA were able to determine that no reservoir communication existed while at stationary conditions. Calculated circulating rates and pressures were established on top of the fill prior to RIH for each sand bite. Based on real time BHA data readings (fig 6) it was confirmed that the ECD generated was not leading to partial losses and full fluid cleanout efficiency could be assumed.5

Figure 5

Well C deviation and completion ID's profile

Figure 5

Well C deviation and completion ID's profile

Figure 6

Base parameter prior RIH

Figure 6

Base parameter prior RIH

Combined BHA Functionality – A Discussion

As previously discussed, a specialized switchable fill cleanout sub was utilized for this job. This sub was designed in a manner that allows exerting max amount of hydraulic jetting force to the fill in the well either in forward or reverse circulating modes by directing 100% of fluid in either the forward or reverse directions. Since flow direction through the tool is the main factor that determines penetration rate and wiper trip efficiency, a poor understanding of the tool functioning downhole can compromise the cleanout. Having the real time data collection system in place made it possible to check the sub status at any given wellbore depth and pump rate, irregardless of surface readout. While function testing the tool there was a clear switch signal (280 Psi) at the BHA while there was only a small change (80 psi) at surface (fig.7).

Figure 7

Switch indication of the circulating sub at the downhole BHA

Figure 7

Switch indication of the circulating sub at the downhole BHA

A pressure drop when flow direction switched also corresponded to expected values (fig.8). The differential pressure reading across the tool was continuously cross checked against the "sub switch calculator" values and a close match concluded the sub was functioning as required (fig. 9, 10) providing further confidence in the cleanout system. Stable pressure drop observed across the BHA at the same pump rates also was an indication of steady functioning of the tool irregardless of what fluids or solids were inside the CT or wellbore (fig.9). It was difficult to make the same assumption based on surface pressure readouts as changing pump pressures could be the result of fluid becoming overloaded by circulating out sand or by a batch of fluid with uneven distribution of friction reducer. It can be observed in (fig.11) that displacing the wellbore volume to slick sea water continuously reduces surface pumping pressure while differential at the BHA stays constant.

Figure 8

Pressure drop calculation at circulating sub switch

Figure 8

Pressure drop calculation at circulating sub switch

Figure 9

Pressure drop at different pump rates

Figure 9

Pressure drop at different pump rates

Figure 10

Pressure drop calculations at different rates

Figure 10

Pressure drop calculations at different rates

Figure 11

Changing of CT pump pressure while circulating wellbore fluid to slick sea water

Figure 11

Changing of CT pump pressure while circulating wellbore fluid to slick sea water

Surface interpretation of wellbore hydraulics

It was discussed that one of the main advantages of a real time downhole communication system in sand cleanout type operations was the ability of monitoring actual downhole pressures at the circulation point. For subject operation monitoring at the circulating point was useful in several ways. Once constant parameters were established, a decision regarding size of bite volume of the sand to be circulated out of hole could be made. Other important data obtained through the electrical CT system during the job was early fluid loss indications (fig.12) indicating the sand plug/bridge holding the weight of fluid above was about to be penetrated and give way. At a very early stage it was identified that small amount of losses start taking place and partial communication with the reservoir was being created. As can be seen from (fig.12) wellbore pressures started fluctuating while performing bites into the sand but soon stabilized. As the cleanout progressed, larger fluid volumes started being lost to the reservoir and when pulling up hole further losses occurred which indicated that no further sand bridges remained. Since the CT operators were aware of the imminent losses appropriate measures were taken in a timely manner to reduce the chance of getting stuck with the CT.

Figure 12

Circulating fluid loss indication at downhole BHA

Figure 12

Circulating fluid loss indication at downhole BHA

Software module predictions versus actual surface and downhole data

Comparison of actual sand cleanout job data versus the software predictions revealed several valuable points First of all it verified the accuracy of the software module.

(Fig.13) is the comparison of surface pump pressure while circulating the CT reel volume with non-slick and slick sea water. When displacing from non-slick to slick sea water at the start of a job it serves as a very good calibration of the fluid friction coefficient and it can be observed that the the fluid coefficient was closely matched to the actual job data across a range of pump rates.

Figure 13

Compariosn of Surface Treating pressures with slick and n/slick sea water while CT in static condition

Figure 13

Compariosn of Surface Treating pressures with slick and n/slick sea water while CT in static condition

For a cleanout type operation, the main advantages of the program are its ability to accurately predict sand mobilization and transportation to surface. This data is based on empirically derived flow loop data. Considering run #1, initially the sand transport model was built assuming no sand in the wellbore to the slickline holdup depth and just a 50 m bite into the fill. This resulted in a prediction of zero sand to surface until the CT BHA reached shallow depth. Importing actual job data (fig.15) into the software and maintaining the same sand distribution yielded results that did not match the surface sand detectors readings. Once sand distribution along the wellbore was introduced into the model above holdup depth the software calculated a similar sand return trend (green line fig.17) that was recorded in reality (fig16). Thus the software was not only helpful in understanding and optimizing the sand cleanout efficiency but also helped the design engineer to understand the picture of initial sand distribution along the wellbore.14

Figure 14

Comparison of Surface treating pressure (actual vs calculated) while CT in dynamic condition

Figure 14

Comparison of Surface treating pressure (actual vs calculated) while CT in dynamic condition

Figure 15

Actual Job parameter, Run #1

Figure 15

Actual Job parameter, Run #1

Figure 16

Clamp-on sand detectors reading, Run #1

Figure 16

Clamp-on sand detectors reading, Run #1

Figure 17

Software calculated sand return rate at surface, Run #1

Figure 17

Software calculated sand return rate at surface, Run #1

It was already discussed that run #2 was only partially successful in cleaning all of the solids during the wiper trip as pumping was incorrectly shut down before CT reached the surface following the first batch of sand cleaning up at the sand detector (fig.19). It was also mentioned that based on downhole pressure tendency at the BHA in comparison to the previous run it was clear that more sand was suspended above the BHA but due to the large size completion could not be unloaded while being stationary with CT during the circulation near surface. While correlating the actual job data (fig.18) to the model (fig.20), the software predicted that 45% solids remained in hole (fig.21). The software could therefore be used to reliably inform the engineer that additional solids remained in the well and that further increases of pump rate or POOH to surface would be required to remove all the sand from the wellbore.

Figure 18

Actual job parameter, Run #2

Figure 18

Actual job parameter, Run #2

Figure 19

Clamp-on sand detectors reading, Run #2

Figure 19

Clamp-on sand detectors reading, Run #2

Figure 20

Software calculated sand return rate at surface, Run #2

Figure 20

Software calculated sand return rate at surface, Run #2

Figure 21

Solid removed versus circulating point with actual job data

Figure 21

Solid removed versus circulating point with actual job data

During run #6 where the BHA was brought all the way to surface a second wave of sand was observed by the sand detectors after the initial wave of sand returns had died off. Importing the actual data from run 6 (fig.23) into the software resulted in the outcome highlighted in (fig.25). To get the model to match with reality, the starting sand volume in the model during run 6 had to be distributed towards the upper end of the deviated section of the completion. It can therefore be concluded that this was the place the large sand particles on previous runs had been dropped to when a full wiper trip to surface had not been performed.2224

Figure 22

Solid removed versus circulating point with pump rate theoretically increased to 4 bpm

Figure 22

Solid removed versus circulating point with pump rate theoretically increased to 4 bpm

Figure 23

Actual job parameter, Run #6

Figure 23

Actual job parameter, Run #6

Figure 24

Clamp-on sand detectors reading, Run #6

Figure 24

Clamp-on sand detectors reading, Run #6

Figure 25

Software calculated sand return rate at surface, Run #6

Figure 25

Software calculated sand return rate at surface, Run #6

Live well deployment system

Before start of actual gun deployment, a dummy trial was performed to identify if any gaps existed in the deployment system and procedures. The result of this was confidence gained that a safe and efficient operation could be performed successfully.

The CT contact friction co-efficienct calculated from previous sand cleanout runs was in the range of 0.29 (fig 27). This value was considered in the software while simulating the required forces to run the perforating BHA to depth. It is worth to note that although calculation was predicting frictional lock before CT reached the target depth, no issues were noted which eliminated the need for contingency steps to reduce wellbore friction. Again having real time CCL data from the BHA provided confidence that the BHA was still moving even after passing into the theoretical friction lock zone.

Figure 26

Surface indication of the gun firing at the Coiled Tubing cab.

Figure 26

Surface indication of the gun firing at the Coiled Tubing cab.

Figure 27

Actual tubing force calculation vs software prediction while CT run in hole

Figure 27

Actual tubing force calculation vs software prediction while CT run in hole

The real time BHA data made it possible to make several correlation passes using CCL and gamma ray before displacing the wellbore from the kill weight fluid used for the sand cleanout to base oil for perforating. The ability of swapping fluids and perforating the well in the same run saved significant time as it eliminated the need for an extra run. Also the fact that electrical firing head was used to activate the gun there was no need for any safety precautions to be taken to escape accidental hydraulic activation of perforators during displacement. Live monitoring of the well also provided actual bottom hole pressures thus exact underbalance conditions were achieved prior to firing the guns.

At the moment of the gun activation, a very clear indication of the gun firing was recorded on surface and downhole. WHP, CTP and CT weight dropped and pressures began to rise equalizing at reservoir pressure (fig 26).

Pulling out of hole with spent guns was carried out without any problem and no overpull was recorded. The reverse deployment process was carried out under live well conditions with 1300 psi wellhead pressure as per the steps described previously. All 22 connectors were recovered from the well easily with only one requiring the retrieval BHA to be slightly modified to release it.

The total CT time on the well for perforating operation was 205 hrs and included; rigging up the surface stack and function testing the same, dummy trial to gain initial experience, running guns into the well and displacing kill fluid to base oil, perforating the well, pulling out of hole and reverse deploying spent guns out of hole and finally rigging down the surface stack. It was calculated that a time saving of four rig days was made when compared alongside performing this job using an e-line tractor.

Considering the length of the perforated interval and complexity of the operation it can be concluded that the selected deployment system offered significant technical advantages over alternative perforating methods. Additionally it reduced operational time, number of steps required and most importantly introduced additional health and safety benefits.

Production Result

Fig 29 shows the well is still producing up to now. An initial test soon after the job (mid-May, 2013) indicated the well was able to produce 4.3 mbopd with 1,038 psi of THP.

Figure 28

Live well deployment stack (perforation job).

Figure 28

Live well deployment stack (perforation job).

Figure 29

Production profile after Well C was handed over to production

Figure 29

Production profile after Well C was handed over to production

Summary

Wrapping up the success story it should be noted that there was a sound engineering approach adopted during the design and execution phases, industry leading, fit for purpose technology deployed and as a result:

  • The well cleanout operation was successfully achieved with one mile of wellbore sand to be cleaned out.

  • Combination of the software module, hydraulic switchable sub and real time downhole communication system increased effectiveness and full understanding of the operation and what is happening downhole and where the initial sand deposition is in the well. Total control was maintained over the equipment during the entire operation and real time data was continuously checked against predicted values.

  • Comparison of the real time data against software predictions confirmed the reliability of the software for sand cleanout job designs. A close match between dynamic and static pressures as well as sand circulating time and solids at surface rates were observed when actual data was laid out alongside the predicted values.

  • Sand arrived at surface in multiple batches, most likely sorted by particle size (and hence mobility) during the wiper trip out of hole and therefore the CT BHA should always be brought to surface during every wiper trip to ensure all sand is transported out of the well. The software also predicted this when the correct initial sand distribution was entered, therefore the software can be used to help the engineer determine the initial sand distribution given the sand recovered back at surface versus time.

  • A simple and reliable live well deployment system made it possible to deploy an extra-long perforating BHA and perforate the target zone in a single run. The system also made it possible to reverse deploy the perforating guns without the need to perform costly and damaging well killing operations. Hydraulic capabilities of the CT saved time as the well was displaced to base oil in the same run as perforating.

  • Having the electrical CT allowed the advantage of using an electrical firing head at the same timeas having the hydraulic firing backup system – almost eliminating the chance of not activating the firing head downhole.

  • It is calculated four rig days were saved during the perforating part of the job alone versus using the historical approach of e-line tractors.1

SI Metric Conversion Factors

bbl ×159 E +00 = liter 
ft × 0.3048 E +00 = m 
inch × 25.4 E–03 = m 
psi × 6.895 E +03 = Pa 
bbl ×159 E +00 = liter 
ft × 0.3048 E +00 = m 
inch × 25.4 E–03 = m 
psi × 6.895 E +03 = Pa 

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Acknowledgement

The authors would like to thank BP and its partner in AIOC, and Baker Hughes for permission to publish this paper. Additonally, recognition goes to the Well Intervention team who flawlessly executed the job.

Nomenclature

    Nomenclature
    AbbreviationExpansion 
  • BHA

    bottom hole assembly

  •  
  • bpm

    barrel per minute

  •  
  • CCL

    casing collar locator

  •  
  • CT

    coiled tubing

  •  
  • CTP

    coiled tubing pressure

  •  
  • HUD

    hold up depth

  •  
  • MD

    measurement depth

  •  
  • OD

    outside diameter

  •  
  • POOH

    pull out of hole

  •  
  • TD

    total depth or target depth

  •  
  • THP

    tubing head pressure

  •  
  • WHP

    wellhead pressure

Taggart
,
M.
,
N.
Murray
and
T.
Sturgeon
.
2011
.
Operational Benefits of Coiled Tubing Enabled with New Real-Time Data Communication System
.
Presented in SPE/ICoTA Coiled Tubing Roundtable
,
Woodlands
,
5–6 April
.
SPE-142714-MS
. http://dx.doi.org/10.2118/142714-MS
Li
,
J.
,
BJ
Lindsey
,
K.
Rahimov
and
S.
Smith
.
2013
.
Combining Tools to Increase Efficiency in Challenging Wellbore Cleanouts
.
Presented in SPE/ICoTA Coiled Tubing & Well Intervention Conference & Exhibition
,
Woodlands
,
26–27 March
.
SPE-163894-MS
. http://dx.doi.org/10.2118/163894-MS
US patent 6,982,008 B2
: "
Coiled Tubing Wellbore Cleanout
",
January
,
2006
.
Li
,
J.
and
G.
Wilde
.
2005
.
Effect of Particle Density and Size on Solids Transport and Hole Cleaning with Coiled Tubing
.
Presented in SPE/ICoTA Coiled Tubing Conference and Exhibition
,
Woodlands
,
12–13 April
.
SPE-94187-MS
. http://dx.doi.org/10.2118/94187-MS
Rahimov
,
Khalid
. and
Smith
,
Simon
.
2004
.
Cleaned Out a Mile of Formation Sands in a Highly Deviated Wellbore with a New Coiled Tubing Real Time Downhole Measurement System
.
Presented in Offshore Techology Conference Asia
,
Kuala Lumpur
,
25-28 March
.
OTC-25008-MS
. http://dx.doi.org/10.4043/25008-MS
Li
,
J.
,
S.
Walker
and
B.
Aitken
.
2002
.
How to Efficiently Remove Sand From Deviated Wellbores With a Solids Transport Simulator and a Coiled Tubing Cleanout Tool
.
Presented in SPE Annual Technical Conference and Exhibition
,
San Antonio
,
29 September - 2 October
.
SPE-77527-MS
. http://dx.doi.org/10.2118/77527-MS
Eldien
,
H. N.
,
M. A.
Al-Anazi
,
R.
Proctor
,
J. B.
Chesson
,
R. M.
Saleh
.
2006
.
Challenging Wellbore Cleanouts with Coiled Tubing Made Easy with Computer Modeling Technology
.
Presented in SPE/ICoTA Coiled Tubing Roundtable
,
Woodlands
,
4–5 April
.
SPE-100129-MS
. http://dx.doi.org/10.2118/100129-MS
Li
,
J.
,
J.
Misselbrook
and
M.
Sach
.
2010
.
Sand Cleanout with Coiled Tubing: Choice of Process, Tools or Fluids
.
Journal of Canadian Petroleum Technology
49
(
08
):
69
-
82
.
SPE-113267-PA
. http://dx.doi.org/10.2118/113267-PA
Li
,
J.
and
S.
Walker
.
2001
.
Sensitivity Analysis of Hole Cleaning Parameters in Directional Wells
.
SPE Journal
6
(
04
):
356
-
363
.
SPE-74710-PA
. http://dx.doi.org/10.2118/74710-PA
Li
,
J.
,
I.
Bayfield
and
G.
Paton
.
2006
.
Effective Heavy Post-Fracturing Proppant Cleanout with Coiled Tubing: Experimental Study and Field Casing History
.
Presented in SPE Annual Technical Conference and Exhibition
,
Dallas
,
9–12 October
.
SPE-101235-MS
. http://dx.doi.org/10.2118/101235-MS
Gilmore
,
T.
,
R.
Leonard
and
S.
Steinback
.
2005
.
Software, Fluids and Downhole Tools for Successful Sand Clean Outs in Any Wellbore Geometry Using Small Coiled Tubing
.
Presented in SPE Annual Technical Conference and Exhibition
,
Dallas
,
9–12 October
.
SPE-97080-MS
. http://dx.doi.org/10.2118/97080-MS
Engel
,
S. P.
and
P.
Rae
.
2002
.
New Methods for Sand Cleanout in Deviated Wellbores Using Small Diameter Coiled Tubing
.
Presented in IADC/SPE Asia Pacific Drilling Technology
,
Jakarta
,
8–11 September
.
SPE-77207-MS
. http://dx.doi.org/10.2118/77204-MS
Walker
,
S.
and
J.
Li
.
2001
.
Coiled-Tubing Wiper Trip Hole Cleaning in Highly Deviated Wellbores
.
Presented in SPE/ICoTA Coiled Tubing Roundtable held in Houston
,
Texas
,
7–8 March
.
SPE-68435-MS
. http://dx.doi.org/10.2118/68435-MS