Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Bob Shoup - Nail Guns Do Not Build Houses


Published on

SIPES Houston Meeting

Published in: Business
  • I have always found it hard to meet the requirements of being a student. Ever since my years of high school, I really have no idea what professors are looking for to give good grades. After some google searching, I found this service ⇒ ⇐ who helped me write my research paper. The final result was amazing, and I highly recommend ⇒ ⇐ to anyone in the same mindset as me.
    Are you sure you want to  Yes  No
    Your message goes here
  • Be the first to like this

Bob Shoup - Nail Guns Do Not Build Houses

  1. 1. 1 You have seen some great technologies and technological innovations here today. Nail guns are also a great technological innovation. They can significantly improve a carpenter’s efficiency. However, if you place a nail gun on a table, nothing happens.
  2. 2. 2 You need a carpenter to use the nail gun to build the house. The carpenter could use a hammer, but the nail gun is more efficient
  3. 3. 3 Workstations are great technological innovations, they can significantly improve an interpreter's efficiency. However, if you place a workstation on a table, nothing happens.
  4. 4. 4 You need a geoscientist to use the workstation to define the subsurface geology to find oil and gas. They could use paper sections and colored pencils, but the workstation is more effective.
  5. 5. However, many geoscientists today rely solely on their workstation to find oil and gas for them. They use the workstation to auto correlate, to highlight amplitudes, AVO, and attribute anomalies, and to make their maps and calculate their reserves.
  6. 6. Here is an example of a map made by a geoscientist who relied on the workstation and not their understanding of geology. The prospect, seen in this investor presentation found on the SEC website, is for a fault propagation fold in the Malataya Basin of Turkey. The map was constructed in the workstation from a grid of 2D data. The mapped horizon is the blue event highlighted with the arrow. 6
  7. 7. Based on this map, would you invest in the prospect? 7
  8. 8. Before you decide to invest, lets look at the Windjammer discovery in the Rovuma Basin of Mozambique. The Windjammer discovery was drilled in a fault propagation fold, similar to that seen in our investment opportunity. 8
  9. 9. The Windjammer well encountered three pay zones; one in the Miocene, a second in the Oligocene, and a third in the Paleocene. The Miocene and Paleocene pays are stratigraphically trapped. The Oligocene pays are structurally trapped in the core of a fault propagation fold. 9
  10. 10. Coming back to our investment opportunity, we can see amplitudes in the core of the fold in line 1, similar to the Oligocene pays seen in the Windjammer well. With Windjammer as an analog, would you invest? 10
  11. 11. Since the prospect consists of a fault propagation fold, perhaps we should look at a fault propagation fold before we decide to invest. This is a Google Earth image of the Sheep Mountain Anticline, a fault propagation fold on the flank of the Big Horn Basin. I have overlain some contours so we can see what the map pattern is for a fault propagation fold. 11
  12. 12. Fault Propagation Folds are asymmetric folds. The back limb of the fold will exhibit a relatively low dip rate that will be constant and will have the same dip as the fault surface, which for Sheep Mountain is between 30 and 45 degrees. The front limb of the fold will exhibit much steeper dip, and may be overturned . The front limb of the Sheep Mountain Anticline is almost 80 degrees. 12
  13. 13. Coming back to our prospect in the Malatay Basin. Look again at the map. Do you see any dip on the front limb of the fold? 13
  14. 14. Whatever contouring algorithm the interpreter used, it did not contour the dip on the front limb of the fold. The map is wrong, and you should not invest, at least until you have re-contoured the map and re-calculated the reserves. 14
  15. 15. There is no question that technology has made our workflow more efficient. Technology has also made it easier for us to see things in the data that we could not see before. But has it really improved our ability to make money? Let s take a moment and go back in time to see how technology has impacted our industries success. 15
  16. 16. We will go back to the early 1980s. In the early 1980 we were pre- email and pre-internet. Computers were capable of running spreadsheets but little else. Workstations at this time consisted of a drafting table, colored pencils and erasers. Our number one exploration tool was our ability to think about the geology we were interpreting. 16
  17. 17. In the early 1980s geoscientists hand-interpreted their seismic, hand- correlated their wells and hand contoured their maps, forcing them to fully understand the geology of their area and their prospect During that time, the industry average exploration success rate was just under 25%. In the Gulf of Mexico, it was ~22%. 17
  18. 18. But in the 1980s, industry began to increasingly use seismic bright spots to define prospects. Now geoscientists need only to find and delineate an anomalous amplitude and drill it. With industry drilling bright spot defined prospects, our exploration success rate jumped up to just over 25%, in other words, no real improvement. 18
  19. 19. What the industry came to learn is that by focusing on seismic bright spots to define prospects, we got away from understanding the geology of our prospects. We also came to learn that bright spots can be caused by many things other than commercial accumulations of oil and gas. 19
  20. 20. In the late 1990s, industry began using Amplitude Versus Offset to define prospects. AVO is a fluid discriminator, so it can help differentiate hydrocarbons from water. Now geoscientists need only to find and delineate an AVO anomaly and drill it. With industry drilling AVO defined prospects, our exploration success rate jumped up to approximately 30%, a slight increase, but hardly the “magic bullet” that many managers had hoped for. 20
  21. 21. We soon came to learn that AVO anomalies can be caused by many things other than commercial accumulations of oil and gas. And, again, by focusing on technology rather than the geology, we drilled many avoidable dry holes, which is the same mistake we made with bright spots. 21
  22. 22. Today, we have seismic inversion and attribute volumes. We have incredibly robust software that allows interpreters to identify porosity and anomalous areas and to map leads and prospects at the push of a few buttons. So with industry drilling prospects defined by the computer, our exploration success rate has now jumped up to 30%, in other words, there has been little to no change in our industry average success rate with technology. 22
  23. 23. I suspect that sooner or later we will come to realize yet again that focusing on technology rather than the geology will cause us to drill unnecessary dry holes. That is because technology does not find oil and gas, interpreters do. Just as nail guns do not build houses, carpenters do. 23
  24. 24. For a geoscientist to be successful, we must make accurate reserve estimates. If we underestimate reserves, then we can cause our company or investor to not drill potentially economic wells. If we overestimate reserves, we can cause our company or investor to drill non-economic wells. Less than 15% of the worlds geoscientists are highly successful, that is, achieve a drilling success rate above the industry average. 24
  25. 25. Those geoscientists that do have above average drilling success rates shared 10 habits which are geologic best practices. See successful-oil-finders/ 25
  26. 26. In a 2005 talk to the Houston Geologic Society Cindy Yeilding (BP) posed the question “Are Workstations Killing Geology?” She noted that our mapping packages allow us to make good looking maps but they are often inaccurate, and that we make bad maps faster than ever. She then pointed out that those inaccurate maps cause us to drill “dumb” wells or unnecessary and avoidable dry holes. She posed several additional questions including regarding our over- reliance on the workstation: Are all geologic views best displayed on a 20” monitor? (NO) Do we spend too much time displaying our interpretation than thinking about it? (YES) Is the philosophy of seismic to simulator flawed? (Absolutely) 26
  27. 27. Moreover, many geoscientists are relying more on the workstation that their own skills and geologic knowledge. Many interpreters today have never hand-contoured a map or made a cross section, nor are they aware of the proper methods or structural models necessary to make valid interpretations. Increased reliance on auto-picking for interpretations delegates the interpretation to the computer.
  28. 28. The reason workstations cause dry holes is that as interpreters become more dependent on the technology they are losing their geologic skills. We are observing that the workstations are doing to geologic skills what the calculator has done to math skills 28
  29. 29. So just how bad are our math skills? Just how bad was demonstrated when a group of high school students were given calculators programmed to give wrong answers for use in a simple math test.
  30. 30. Almost all of the students turned in tests with all wrong answers, mostly quite obviously wrong. The students were asked if they noticed anything wrong? There answer was yes, but since the calculator said this was the answer, that it must be, even though they ‘felt’ that the answer was wrong.
  31. 31. Our comfort with the workstation has made us complacent, and we now tend to accept what comes out of the workstation without question, even when it is wrong.
  32. 32. 32 So it is important for interpreters to use the workstations as a tool to interpret the data. For those of you who rely on the workstation to interpret the data, you need to be aware of how the computer sees the data
  33. 33. 33 When geoscientists look at this line, they should see a rift basin with some minor inversion. As a geoscientist, you can develop a familiarity with rift basins which you can use to predict where the source rocks were deposited, where the reservoirs were deposited, and where hydrocarbons will migrate
  34. 34. 34 When a computer looks at this line, it sees digits, So a computer can not make geologic interpretations, it makes statistical correlations. A computer cannot develop a familiarity with rift basins nor can it predict anything.
  35. 35. 35 Before a carpenter builds a house with a nail gun, he needs to know what a house looks like and he should know how to build it if he only had a hammer
  36. 36. 36 Likewise, before a geoscientist uses a workstations to interpret a prospect, they should know the geology of the area that they are interpreting. They should also know all of the methods and techniques needed to fully evaluate the subsurface and to make accurate maps.
  37. 37. Lets look at a case study where reserves were added to a field simply by the construction of a geologic cross section. 37
  38. 38. 38 The 9300 Foot Sand in a field producing from multiple reservoirs was an under-producing reservoir. The 9300 Foot Sand was completed in well #10 and produced about 12,000 BO. The well was estimated to have been able to produce over 500,000 BO. Well #4 produced gas and rapidly pressure depleted. The 9300 Foot Sand was abandoned for several years until a new study of the field has done. It was recognized that based on this map that there was a material balance problem for the 9300 Foot Sand. It was also recognized by the young geologist doing the study that no cross section had been constructed for this reservoir.
  39. 39. 39 In doing the new study the geologist constructed a correlation cross section to study the reservoir in greater detail. You can see that wells 3 and 4 encounter a thick channel sand. Wells 7 and 10 encounter overbank or crevasse splay deposits The low production from well 10 was from the crevasse splay sands. The gas production was from the transgressive sand capping the sequences. The main channel sand had not been produced.
  40. 40. 40 Now that the review team had a new depositional model, they constructed a Log Facies map from the regional wells. We can see from that map that the field was situated in one branch of the channel system
  41. 41. 41 The review team then constructed a net pay isochore for the channel sand and recognized two possible channels in the field, one in the central portion of the field, and another on the western margin of the field.
  42. 42. 42 The team then constructed a net pay isochore for the channel sand. The volume potential in the western branch of the channel were too small to develop. The central channel was determined to have approximately 4 million barrels of recoverable reserves remaining. Well 4 was re-completed in the channel sand and found near-virgin pressures.
  43. 43. That example is not the only example of finding additional reserves by constructing a cross section. 43
  44. 44. 44 In this case, the operator had produced this field for a number of years, recovering ~ 1.5 MMBO from the oil rim. They were now ready to blow down the gas cap. Before granting the operator approval to blow down the gas cap, the regulatory agency wanted proof that the oil rim had been fully produced. In compliance with the regulatory agency's request, the operator conducted a complete geological study of the reservoir, including construction of a correlation cross section
  45. 45. 45 The correlation Cross section showed that there was a thick incised valley sand in wells 12, 17, and 3. Wells 12 and 17 were in the gas cap and well 3 was below the water level Incised valley sequences are rarely in communication with the sequence they incise so it was considered unlikely that this incised valley sequence had been produced.
  46. 46. 46 The review team developed 2 different incised valley interpretations. With interpretation 1, the incised valley crossed the field from northeast to southwest and had a volume potential of 3.5 MMBO Two locations were proposed for testing this interpretation. If the first well encountered the incised valley, they would drill the second proposed well.
  47. 47. 47 In the second interpretation, the incised valley missed the oil rim. However, the re- mapping showed that there was an additional volume potential of 1.5 MMBO remaining in the oil rim. If the first well did not encounter the incised valley sequence then they would not drill the second of the two proposed wells.
  48. 48. 48 The first of the two proposed wells encountered the incised valley sand as predicted in interpretation 1, so the second proposed well was drilled The incised valley sequence was found to be thicker than originally thought such that the final reserve add was 4.5MMBO
  49. 49. If you are not constructing cross sections you do not understand the distribution of the fluids or the nature of the reservoir. If you do not understand those, you are most likely leaving reserves behind. Looking at this cross section, one can see that in the uppermost sand of the eastern fault block the perforations (purple rectangles near the depth track) are near the water level, leaving significant attic reserves. 49
  50. 50. The time to construct a cross section and interpret varies from a few hours to a few days, depending on how complex the geology is. Finding additional reserves with that cross section is priceless. 50
  51. 51. We will look at another best practice, constructing a fault surface map and integrating that map with the horizon map to define the fault traces in their proper position 51
  52. 52. This field has been producing for several years, at which time the company sold the field along with several other assets. The field, mapped with 3D seismic and well control, consists of a three-way fault closure against a large growth fault. In addition to the producing fault block, a prospective fault block had been identified and mapped to the north of the producing fault block. The original interpreters had not made a fault surface map. They had simply picked faults on the seismic and posted a fault polygon. 52
  53. 53. Since the original operator had never made a fault surface map, one of the first things the new operator did following the purchase of the field was to construct a fault surface map and integrate it with the horizon map. The fault surface map was constructed from the 3D data as well as the fault picks in the wells. The data posted by the wells is the amount of missing section (Vertical Separation) observed in the wells and the depth of the fault. 53
  54. 54. When the producing horizon was integrated with the fault surface map, the position of the fault traces shifted to the east, opening up approximately 400 acres of additional closure; adding almost 20 BCF of recoverable reserves. 54
  55. 55. In addition to finding additional reserves by constructing fault surface maps, we can also use them to avoid drilling dry holes. 55
  56. 56. 56 Here is a structure contour map generated with 2D seismic. Is it geometrically valid? Since we are dealing with a 2 dimensional map that is portraying a 3 dimensional surface, it takes more understanding, investigation, and analysis to identify geometric problems.
  57. 57. 57 The prospect was drilled and the result was a dry hole Let’s examine the trapping fault: The amount of offset along the trapping fault decreases from east to west. The offset then increases again but the sense of throw is to the north, or up-thrown direction. There is almost no offset at the well location This is a screw fault. Screw faults CANNOT occur in compressional or extensional faults (Scissors fault can occur with strike-slip faults). Therefore, the map portrays a structure that is geometrically, and geologically, impossible.
  58. 58. 58 We call fault interpretations like this “Screw Faults” since the change in the direction of the direction of offset of the fault makes it appear as if the fault is screwing itself through the section. We also call the screw faults because if you drill a well in a prospect set up by a screw fault, you are screwed.
  59. 59. 59 Screw Faults indicate 2 faults have been mapped as 1 fault. This is a result of the interpreters picking fault sticks and using them to post a fault polygon as opposed to mapping the fault. The failure to map the fault surface results in the company drilling an unnecessary and avoidable dry hole.
  60. 60. Screw fault interpretations are very common in our industry. So common that we can see a screw fault in the Petrel User Manual.
  61. 61. If the interpreter had constructed a fault surface map it would have been immediately apparent that the interpretation of the fault as a single fault was geometrically and geologically impossible. A fault surface map takes only a few hours to construct. The failure to map the fault surface cost this company $5.5 million dollars. 61
  62. 62. Technology makes our workflow much more efficient. And, it can help us to find and better develop oil and gas. 62
  63. 63. But ONLY, if we use it as a tool to help us understand the geology of our prospects and fields. If we use technology to replace our thinking and our knowledge of the petroleum system, we will continue to drill unnecessary and avoidable dry holes. 63
  64. 64. For no matter how robust technology becomes, oil is first found in the mind. Technology is nothing more than a tool to help the mind find it. And nail guns are just a tool to help carpenters build houses 64