Society of Petroleum Engineers Distinguished Lecturer Program spe

3. OUTLINE OF PRESENTATION INTRODUCTION CLASSIC APPROACH TO SAND CONTROL ANALYSIS OF FINES AND CONSEQUENCES CASE HISTORIES RECENT ADVANCES IN FINES STABILIZATION LESSONS LEARNED AND BEST PRACTICES

4. The idea here is to start the talk by pointing out that different parts of the total production train sees fines production in a different manner. The completions engineers see it as a matter of controlling �sand� and not really fines. If we control fines, we decrease production. The �operations� engineers see fines production as something to pass on to the facilities engineers. The �flow assurance� engineers see fines as a threat to moving fluids from point A to B, and the problems associated with pigging, cost of pipelines, etc. The facilities engineers really do not have a clue! Their whole purpose in life is to accept the �fines� that the formation and completion engineers give them, do something with the fines ( as little as possible ) and pass on to someone else the �fines� problem. The whole aim of the talk is to show how and why each discipline needs to talk to each other and make dealing with fines issue an holistic approach.The idea here is to start the talk by pointing out that different parts of the total production train sees fines production in a different manner. The completions engineers see it as a matter of controlling �sand� and not really fines. If we control fines, we decrease production. The �operations� engineers see fines production as something to pass on to the facilities engineers. The �flow assurance� engineers see fines as a threat to moving fluids from point A to B, and the problems associated with pigging, cost of pipelines, etc. The facilities engineers really do not have a clue! Their whole purpose in life is to accept the �fines� that the formation and completion engineers give them, do something with the fines ( as little as possible ) and pass on to someone else the �fines� problem. The whole aim of the talk is to show how and why each discipline needs to talk to each other and make dealing with fines issue an holistic approach.

5. FINES PRODUCTION CAN BE A BIG PROBLEM !! This slide is to illustrate that fines production can really be a problem! This slide is from a separator in Sumatra where we do good �sand control� with gravel packs, but still produce tonnes of fines through the sand control.This slide is to illustrate that fines production can really be a problem! This slide is from a separator in Sumatra where we do good �sand control� with gravel packs, but still produce tonnes of fines through the sand control.

6. FINES PRODUCTION CAN BE A REAL BIG PROBLEM !! This slide is just another example of where we have good sand control in place in Sumatra and make tonnes of fines. This slide is a side view of the previous slide.This slide is just another example of where we have good sand control in place in Sumatra and make tonnes of fines. This slide is a side view of the previous slide.

7. IS THE CLASSIC SAND CONTROL APPROACH ENOUGH ?? THE CLASSIC SAND CONTROL APPROACH Obtain sample of core material Measure particle size distribution Select screen size using 1 x D10 Select gravel size using 6 x D50 REALLY FAIRLY STRAIGHT FORWARD�.RIGHT ??

8. NORTH SEA FORMATION This is an example of a Chevron field in the North Sea where the classic approach to selecting the gravel pack sand and sand control screen was used. The PSD of this sample shows the formation contains ~ 13% sub 44 micron material,and has a D10 of about 200 microns. At the time the wells were completed, fines migration tests were not commonly used, and it was assumed that the sand control used in the wells was good enough to control any �sand�. Based on commonly accepted practices, this well was completed horizontally using a 20/40 gravel pack and a 220 micron premium screen. All went well until water hit. The following two slides show a portion of the production history of two wells in the field. This is an example of a Chevron field in the North Sea where the classic approach to selecting the gravel pack sand and sand control screen was used. The PSD of this sample shows the formation contains ~ 13% sub 44 micron material,and has a D10 of about 200 microns. At the time the wells were completed, fines migration tests were not commonly used, and it was assumed that the sand control used in the wells was good enough to control any �sand�. Based on commonly accepted practices, this well was completed horizontally using a 20/40 gravel pack and a 220 micron premium screen. All went well until water hit. The following two slides show a portion of the production history of two wells in the field.

9. FINES PRODUCTION � NORTH SEA WELL Inspection of this production history clearly shows that at some point as the water cut came up, �sand� started to be produced. A PSD of the �sand� produced and bailed from the well showed that the majority of the �sand� had an average particle size of ~ 50 � 80 microns. This well was choked back to minimize the production of the �sand�Inspection of this production history clearly shows that at some point as the water cut came up, �sand� started to be produced. A PSD of the �sand� produced and bailed from the well showed that the majority of the �sand� had an average particle size of ~ 50 � 80 microns. This well was choked back to minimize the production of the �sand�

10. HOW DO YOU KNOW IF FINES MIGHT BE A PROBLEM FOR YOU?? The purpose of this slide is to point out that there are two classic different ways to do an analysis of solids in a formation. The classic approach of using a sieve analysis is one way, and the laser particle size analysis is another. The two techniques will give different results. There is a difference because of the way the samples are prepared and the method the samples are analyzed. The dry or wet sieve technique requires breaking up the sample to some degree with a mortor and pestle just like the LPSA technique. The LPSA technique goes another step further by requiring the sample be �sonicated� to further break up the sample. This sonication step separates the �fines� from the rest of the sample. The main difference is that the LPSA technique shows ALL the fines; where as, the sieve technique does not. Therefore, the sieve technique is very conservative with respect to fines; where as, the LPSA technique is very pessimistic with respect to fines. Why is this? The next slide show why��..The purpose of this slide is to point out that there are two classic different ways to do an analysis of solids in a formation. The classic approach of using a sieve analysis is one way, and the laser particle size analysis is another. The two techniques will give different results. There is a difference because of the way the samples are prepared and the method the samples are analyzed. The dry or wet sieve technique requires breaking up the sample to some degree with a mortor and pestle just like the LPSA technique. The LPSA technique goes another step further by requiring the sample be �sonicated� to further break up the sample. This sonication step separates the �fines� from the rest of the sample. The main difference is that the LPSA technique shows ALL the fines; where as, the sieve technique does not. Therefore, the sieve technique is very conservative with respect to fines; where as, the LPSA technique is very pessimistic with respect to fines. Why is this? The next slide show why��..

11. THE DIFFERENCE BETWEEN LASER AND SIEVE ANALYSES Here we have cartoon examples of a LPSA and sieve analysis. The LPSA technique uses a slurry of particles that reflect the light of a laser beam. Because the particles are �moving� in the slurry, the laser beam sees an �average� particle size equal to the longest average axis of the particle as shown in the top cartoon. This causes the average particle size to �look� bigger by giving an �aspect� ratio of �1� to every particle. The sieve method on the other hand is a technique that is based on a �weight� of solids trapped on a sieve. The method relies on the particles �bouncing� around ( in the case of the dry sieve method) on a specific sieve size. As depicted in the cartoon, it is then possible for particles with more weight to pass through a specific sieve size to give a large �weight� to a smaller sieve size. This can result in the distribution showing more small particles than is really there. Therefore, comparing the two techniques, it would seem that the LPSA technique would show more larger particles than the sieve method; however, the LPSA method would show ALL the fines. The following slide is an example of just this effect.Here we have cartoon examples of a LPSA and sieve analysis. The LPSA technique uses a slurry of particles that reflect the light of a laser beam. Because the particles are �moving� in the slurry, the laser beam sees an �average� particle size equal to the longest average axis of the particle as shown in the top cartoon. This causes the average particle size to �look� bigger by giving an �aspect� ratio of �1� to every particle. The sieve method on the other hand is a technique that is based on a �weight� of solids trapped on a sieve. The method relies on the particles �bouncing� around ( in the case of the dry sieve method) on a specific sieve size. As depicted in the cartoon, it is then possible for particles with more weight to pass through a specific sieve size to give a large �weight� to a smaller sieve size. This can result in the distribution showing more small particles than is really there. Therefore, comparing the two techniques, it would seem that the LPSA technique would show more larger particles than the sieve method; however, the LPSA method would show ALL the fines. The following slide is an example of just this effect.

12. COMPARISON OF RESULTS This slide is some actual data from one of Chevron�s wells where we did both methods. The consequence of using either technique is obvious. If you use the sieve data for sizing a sand control screen (i.e., the D10) the screen opening is much smaller than the opening of the screen as compared to if you use the LPSA data. Also, just looking at the sieve data, you would think there is little possible problem with the potential for fines production. BUT WAIT�there is more to the story !!!!!!This slide is some actual data from one of Chevron�s wells where we did both methods. The consequence of using either technique is obvious. If you use the sieve data for sizing a sand control screen (i.e., the D10) the screen opening is much smaller than the opening of the screen as compared to if you use the LPSA data. Also, just looking at the sieve data, you would think there is little possible problem with the potential for fines production. BUT WAIT�there is more to the story !!!!!!

13. WHAT IF THERE IS A LARGE AMOUNT OF FINES ?

14. GULF OF MEXICO FORMATION It is important to keep in mind that the objective of the lecture is to educate the listener about the impact of fines from the formation on the completion, the facilities and related situations like water injection wells. With this in mind, the slides presented here are a sample of the initial part of the presentation that is to educate the listener about how to evaluate if fines are a potential problem. This slide is a PSD of a formation in the GOM where we were going to frac pack the well, but there was concern about the potential for fines production after the completion. Just looking at the PSD, an engineer might be very concerned since there is from 9 � 19% sub 44 micron particles in the formation. The GOM facility engineers classically do not put much if any equipment on the topsides for handling much �sand�; i.e., fines. Therefore, if the fines are prone to move, the sand and screen for the frac pack would have to be sized to help control some amount of the fines. This would potentially result in a finer gravel and a smaller micron rating for the premium sand control screen. This in turn might result in an unfavorable pressure drop across the completion at the high rates expected for the well. The only way to learn more about the potential for fines migration was to do a fines migration test; i.e., a Critical Rate Test (CRT). The next slide shows the results of the CRT. It is important to keep in mind that the objective of the lecture is to educate the listener about the impact of fines from the formation on the completion, the facilities and related situations like water injection wells. With this in mind, the slides presented here are a sample of the initial part of the presentation that is to educate the listener about how to evaluate if fines are a potential problem. This slide is a PSD of a formation in the GOM where we were going to frac pack the well, but there was concern about the potential for fines production after the completion. Just looking at the PSD, an engineer might be very concerned since there is from 9 � 19% sub 44 micron particles in the formation. The GOM facility engineers classically do not put much if any equipment on the topsides for handling much �sand�; i.e., fines. Therefore, if the fines are prone to move, the sand and screen for the frac pack would have to be sized to help control some amount of the fines. This would potentially result in a finer gravel and a smaller micron rating for the premium sand control screen. This in turn might result in an unfavorable pressure drop across the completion at the high rates expected for the well. The only way to learn more about the potential for fines migration was to do a fines migration test; i.e., a Critical Rate Test (CRT). The next slide shows the results of the CRT.

15. FINES MIGRATION TEST � GOM FORMATION This graph is the data from the Critical Rate Test (CRT). The test was done with brine, which is the worst possible situation since the formation is water wet. The results of the CRT show that up to the maximum rate expected, there was no evidence of fines movement. Considering that there is 9 � 19% sub 44 micron material in the formation, it was not easily understood why the fines did not move. Further investigation was done using thin sections and SEM analyses. The next slide is a thin section of the formation sand.This graph is the data from the Critical Rate Test (CRT). The test was done with brine, which is the worst possible situation since the formation is water wet. The results of the CRT show that up to the maximum rate expected, there was no evidence of fines movement. Considering that there is 9 � 19% sub 44 micron material in the formation, it was not easily understood why the fines did not move. Further investigation was done using thin sections and SEM analyses. The next slide is a thin section of the formation sand.

16. THIN SECTION OF GOM FORMATION Evaluation of the thin section indicates that there are large pore throats and there are not many fines in the pore throats. This helps, but still does not explain why there is 9 � 19% sub 44 micron material in the PSD analysis yet there is no evidence of fines movement. The SEM tells the rest of the story. Turn this around�. Doing the PSD is not near enough to understandEvaluation of the thin section indicates that there are large pore throats and there are not many fines in the pore throats. This helps, but still does not explain why there is 9 � 19% sub 44 micron material in the PSD analysis yet there is no evidence of fines movement. The SEM tells the rest of the story. Turn this around�. Doing the PSD is not near enough to understand

17. SEM OF GOM FORMATION The SEM shows that the fines are mostly kaolinite and are not cementing fines, but are located in pockets around the smaller pore throats. It was then reasoned that the fines did not move because the preferred flow path for the fluids is through the larger pore throats and therefore the fines in the smaller pore throats were not subjected to the high fluid flow rates. The consequence of these findings is that the completion was able to use 16/30 gravel instead of 20/40 and the screen mesh size was increased to the next higher size. The wells are now producing 15,000 � 20,000 BOPD with < 10% water cut and there is no evidence of any solids production. The SEM shows that the fines are mostly kaolinite and are not cementing fines, but are located in pockets around the smaller pore throats. It was then reasoned that the fines did not move because the preferred flow path for the fluids is through the larger pore throats and therefore the fines in the smaller pore throats were not subjected to the high fluid flow rates. The consequence of these findings is that the completion was able to use 16/30 gravel instead of 20/40 and the screen mesh size was increased to the next higher size. The wells are now producing 15,000 � 20,000 BOPD with < 10% water cut and there is no evidence of any solids production.

18. CONSEQUENCE OF FINES MIGRATION TEST RESULTS Larger gravel used in frac pack Larger screens used in completion Topsides Installed minimal sand handling equipment Saved several million $$�S WELL ON PRODUCTION FOR ~ 5 YEARS AT EXPECTED OIL PRODUCTION RATE; MAKING WATER AND NO SAND

19. The obvious consequences of fines/solids production are cut screens, erosion of downhole equipment, filling the wellbore, filled separators, and emulsions. However, in today�s world where we are going to more and more zero discharge, produced water re-injection is becoming a big issue. In many cases, the facilities engineers set specs for produced water coming from the facilities based on not what the formation requires, but what is convenient for the facilities to provide. A common set of spec from the facilities is 50 micron solids at concentrations ranging from 5 ppm � 25 ppm. This has a very big impact on the completion of the injection well. The next example is a case Chevron is currently studying and the impact of the solids is very significant !! For one of Chevron�s fields, the sand in the productive intervals has a UCS that ranges from 800 � 200 psi. The BOD for the production wells is a sub-sea cased hole frac pack using 20/40 gravel with a 250 micron premium sand control screen. The BOD of the sub-sea injection wells is a cased hole frac pack with 16/30 gravel. The project requires injection of 60 � 75 kbwpd per well, and facilities is saying the water quality will be 50 micron solids at a concentration of 10 � 14 ppm IF solids are produced with the produced water. The obvious consequences of fines/solids production are cut screens, erosion of downhole equipment, filling the wellbore, filled separators, and emulsions. However, in today�s world where we are going to more and more zero discharge, produced water re-injection is becoming a big issue. In many cases, the facilities engineers set specs for produced water coming from the facilities based on not what the formation requires, but what is convenient for the facilities to provide. A common set of spec from the facilities is 50 micron solids at concentrations ranging from 5 ppm � 25 ppm. This has a very big impact on the completion of the injection well. The next example is a case Chevron is currently studying and the impact of the solids is very significant !! For one of Chevron�s fields, the sand in the productive intervals has a UCS that ranges from 800 � 200 psi. The BOD for the production wells is a sub-sea cased hole frac pack using 20/40 gravel with a 250 micron premium sand control screen. The BOD of the sub-sea injection wells is a cased hole frac pack with 16/30 gravel. The project requires injection of 60 � 75 kbwpd per well, and facilities is saying the water quality will be 50 micron solids at a concentration of 10 � 14 ppm IF solids are produced with the produced water.

20. WEST AFRICA DEEPWATER FORMATION The PLSA indicates up to 20% sub-micron particlesThe PLSA indicates up to 20% sub-micron particles

21. THIN SECTION OF WEST AFRICA FORMATION The thin section analysis shows lots of fine loose particles that are not necessarily clays. The question is�..do the fines move, and if fines move what are the fines that move, and how much will move? A fines migration test was run. The test did not show any change in permeability of the core material��BUT�..The thin section analysis shows lots of fine loose particles that are not necessarily clays. The question is�..do the fines move, and if fines move what are the fines that move, and how much will move? A fines migration test was run. The test did not show any change in permeability of the core material��BUT�..

22. SOLIDS PRODUCED IN FINES MIGRATION TEST Fines were produced out of the core at a very low interstitial brine flow rate. If one inspects the scale on the photo, it is obvious that some of the fines are less than 40 microns and small enough to pass through a 20/40 gravel pack. Since the BOD of the production wells is a cased hole frac pack using 20/40 gravel and a 250 micron premium screen, it is very obvious that fines will be produced from the wells. From the specs of the facilities engineers, the fines that are produced through the gravel pack will be passed on through the facilities to be re-injected. There is a real question then�.is it possible to maintain injection of 60 � 75 kbwpd of produced containing 50 micron and smaller particles into a sub-sea cased hole frac pack completion for the life of the field? Fines were produced out of the core at a very low interstitial brine flow rate. If one inspects the scale on the photo, it is obvious that some of the fines are less than 40 microns and small enough to pass through a 20/40 gravel pack. Since the BOD of the production wells is a cased hole frac pack using 20/40 gravel and a 250 micron premium screen, it is very obvious that fines will be produced from the wells. From the specs of the facilities engineers, the fines that are produced through the gravel pack will be passed on through the facilities to be re-injected. There is a real question then�.is it possible to maintain injection of 60 � 75 kbwpd of produced containing 50 micron and smaller particles into a sub-sea cased hole frac pack completion for the life of the field?

23. CONSEQUENCE OF FINES PRODUCTION The water injection program for each injector in the field is 40 � 65 kbwpd for 10 to 12 years. The BOD for these wells is a CHFP, but these results show that injectivity can not be maintained for the desired life of the well with a CHFP completion. A cased hole perforated completion is out of the question since water hammer studies indicate that even 1 emergency shut-in will cause severe sand production. To address the problem, the facility engineers are going to have to design the topsides to do filtration much lower than the BOD of 50 microns and 10 lb/1000 bbls. This will add a great expense to the cost of the project. The water injection program for each injector in the field is 40 � 65 kbwpd for 10 to 12 years. The BOD for these wells is a CHFP, but these results show that injectivity can not be maintained for the desired life of the well with a CHFP completion. A cased hole perforated completion is out of the question since water hammer studies indicate that even 1 emergency shut-in will cause severe sand production. To address the problem, the facility engineers are going to have to design the topsides to do filtration much lower than the BOD of 50 microns and 10 lb/1000 bbls. This will add a great expense to the cost of the project.

24. SAND DEPOSITION IN FLOWLINES Pressure losses due to reduced flow area Pigging requirement to remove sand Under sand bed corrosion pitting One of the biggest problems associated with flow assurance issues is that there is not a good model for solids transport in flow lines. The University of Tulsa and some operators have been working on such models. A good example of the importance of understanding flow assurance issues and fines production is a major gas field that Chevron is in Phase II of developing. One of our partners did fines migration tests and told us that we should expect ~ 0.22 lbs fines produced with every MMSFG. Well these wells are expected to make ~ 350,000 MMSCFGPD each. There are potentially 6 subsea wells flowing to a common pipeline to shore. This means about 450 lbs solids per day from the central gathering location. If these figures are correct, it means potentially having to re-route the pipeline to minimize dooning in the pipeline ( at a cost of ~ $200 MM), the pipeline may have to be resized from 40� to 42� at a cost of $42MM. This does not take into account the added cost of pig launchers, addition of corrosion inhibitors because of under bed corrosion, special engineering to take into account potential erosion issues, etc. For this project, fines production is potentially VERY expensive.One of the biggest problems associated with flow assurance issues is that there is not a good model for solids transport in flow lines. The University of Tulsa and some operators have been working on such models. A good example of the importance of understanding flow assurance issues and fines production is a major gas field that Chevron is in Phase II of developing. One of our partners did fines migration tests and told us that we should expect ~ 0.22 lbs fines produced with every MMSFG. Well these wells are expected to make ~ 350,000 MMSCFGPD each. There are potentially 6 subsea wells flowing to a common pipeline to shore. This means about 450 lbs solids per day from the central gathering location. If these figures are correct, it means potentially having to re-route the pipeline to minimize dooning in the pipeline ( at a cost of ~ $200 MM), the pipeline may have to be resized from 40� to 42� at a cost of $42MM. This does not take into account the added cost of pig launchers, addition of corrosion inhibitors because of under bed corrosion, special engineering to take into account potential erosion issues, etc. For this project, fines production is potentially VERY expensive.

25. ORGANIC POLYMER STABILIZATION In theory, the silicate surface is coated with a very thin layer of the chemical. The chemical is polymeric in nature with small �tails� �hanging off the surface of the silicate and the �fines�. The binding of the two polymer coated rock surfaces is a result of the Van der Waal forces between the polymeric tails. It is the attraction of these �tails� that �bind� the fines to the surface of the rock. The next slide shows a SEM of a rock treated with the polymeric material. There is no obvious signs of the polymer on the surface of the rock.In theory, the silicate surface is coated with a very thin layer of the chemical. The chemical is polymeric in nature with small �tails� �hanging off the surface of the silicate and the �fines�. The binding of the two polymer coated rock surfaces is a result of the Van der Waal forces between the polymeric tails. It is the attraction of these �tails� that �bind� the fines to the surface of the rock. The next slide shows a SEM of a rock treated with the polymeric material. There is no obvious signs of the polymer on the surface of the rock.

26. ORGANIC POLYMER STABILIZATION This is a SEM of a core treated with the new polymeric system. It does not appear that the treatment has affected the apparent perm of the core at all. It is not even obvious that the core has been treated.This is a SEM of a core treated with the new polymeric system. It does not appear that the treatment has affected the apparent perm of the core at all. It is not even obvious that the core has been treated.

27. INORGANIC POLYMER STABILIZATION Another approach is to use an organosilane system. This system is dissolved in diesel and applied in a series of flushes. The system is not designed to �consolidate� the formation. It is designed to add enough strength and binding to keep solids and fines in place. The strength needed is determined based on the projected production rates.Another approach is to use an organosilane system. This system is dissolved in diesel and applied in a series of flushes. The system is not designed to �consolidate� the formation. It is designed to add enough strength and binding to keep solids and fines in place. The strength needed is determined based on the projected production rates.

28. INORGANIC POLYMER STABILIZATION This is a case history of one well treated with the product. The treatment was done immediately after the shut-in and lasted for a few months. However, it is seen that solids production restarted again. This products is still in the development stage, but one operator has treated 8 wells. This has included multiple treatments in some cases.This is a case history of one well treated with the product. The treatment was done immediately after the shut-in and lasted for a few months. However, it is seen that solids production restarted again. This products is still in the development stage, but one operator has treated 8 wells. This has included multiple treatments in some cases.

29. LESSONS LEARNED & BEST PRACTICES Sieve & laser methods of particle size analyses can show different results Use PSD�s, thin sections & SEM�s to screen for potential fines concerns Fines migration tests are needed to determine if fines actually move

30. LESSONS LEARNED & BEST PRACTICES Some consequences of fines production are not so obvious Sand bed corrosion Sand control design for water injectors New approaches to fines stabilization include organosilanes & organic polymers There is a great need for a good solids transport model