Unconventional oil production in the United States is growing. Liquid rich shale plays are relatively new, making it difficult to estimate future production. Each play is different, requiring a unique well design. Additional difficulties are seen within each basin as geology can change from one mile to the next. It is important to consider each area individually before modeling production. EOG Resources (EOG) is a Bakken top producer. It has developed most of Parshall Field, with only a handful of results from Statoil (STO), Whiting (WLL), and Hess (HES). By using EOG, I was able to isolate results from wells completed in 2007. This provided enough information to use historical data to model EURs for Parshall Field.
Horizontal wells share like characteristics. High initial production or IP rates are followed by high depletion rates. IP rates are used to calculate estimated ultimate recoveries or EURs. The method of calculation is important as this can cause a wide range of results. More importantly, models can be used to produce incorrect results by changing data. This can be from ignorance, but some will manipulate data to produce findings to back their assertions. This causes contradictions difficult to disprove. Different areas see different types of production further complicating well results. Producers also have better or worse results depending on well design.
Well depletion changes from year to year. Models calculate for this, but these are just estimates. It does not account for water, proppant, choke usage. Water and proppant can reduce depletion holding open fractures in the source rock. Ceramic proppant does this better and for a longer period of time. A tight choke will produce a lower IP rate for as long as 30 days. This may or may not affect EURs, but it does change early depletion. Initially, wells experience a hyperbolic initial decline. This decline occurs after a large amount of resource is produced from the stimulated source rock. Depletion continues to decrease until terminal decline begins. This is a change as the shale's matrix replaces production from the stimulated rock. In the Bakken, terminal decline is 8% and continues for decades.
Models are currently used because liquids rich unconventional shale plays have not produced long enough to provide proper historical data. This is further complicated by the number of stages, choke, amount and type of proppant. Some variables are difficult to document. An operator's skill in drilling and completion can only be estimated by current results. Some areas have seen enough development to derive a sample, and use historical production to model EURs. Two important factors to isolate are the area (geology) and operator. Geology can affect production mix, depletion, and total production. Operator skill in drilling and completion methods are also important. I chose EOG Resources based on its consistent outperformance in the Bakken and other U.S. plays. Parshall Field has been its focus, so I used this area to provide well data. Parshall Field has a significant number of well results going back to 2007. This provides five years of production. This time frame sees the largest variation in production. After this time, terminal decline can be used to measure future production.
EOG Well Depletion In Parshall Field In Bo/d
Table 1
| Well | IP Rate | 360- Day IP | 720-Day IP | 1080-Day IP | 1440-Day IP | 1800-Day IP |
| 16637 | 970 | 449 | 313 | 251 | 215 | 234 |
| 16461 | 1487 | 469 | 410 | 321 | 264 | 231 |
| 16543 | 1015 | 440 | 321 | 261 | 226 | 197 |
| 16532 | 1285 | 489 | 365 | 291 | 253 | 223 |
| 16467 | 783 | 333 | 240 | 198 | 171 | 150 |
| 16497 | 1675 | 502 | 364 | 292 | 246 | |
| 16635 | 581 | 286 | 203 | 162 | 135 | 116 |
| 16484 | 1670 | 535 | 379 | 302 | 254 | 219 |
| 16457 | 922 | 481 | 339 | 275 | 232 | 202 |
| 16550 | 1329 | 421 | 307 | 253 | 217 | |
| 16671 | 1198 | 466 | 340 | 282 | 242 | 212 |
| 16483 | 870 | 494 | 396 | 321 | 272 | 239 |
| 16370 | 1553 | 421 | 332 | 269 | 226 | 195 |
| 16469 | 1267 | 533 | 396 | 322 | 276 | 249 |
| 16578 | 817 | 489 | 358 | 291 | 253 | |
| 16371 | 918 | 552 | 421 | 342 | 288 | 250 |
| 16713 | 781 | 740 | 494 | 389 | 322 | |
| 16768 | 1441 | 793 | 535 | 424 | 350 | |
| 16795 | 1519 | 814 | 561 | 438 | 360 | |
| Avg. | 1162 | 511 | 372 | 299 | 253 | 209 |
Rate | 56% | 27.3% | 19.6% | 15.4% | 17.4% |
Table 1 provides how EOG's early wells deplete. Some of its wells had some production problems. These were minor, so I was able to use all 19 Parshall Field wells completed in 2007. My estimates for this time frame has EOG using roughly twice the water and proppant of other operators at that time. Every well is short lateral, so doubling the EURs is required for a proper comparison to current results. The first year of production produced a lower depletion than I would have expected. EOG's use of tight chokes are probably the reason, as well pressures are maintained for a longer period of time.
2007 EOG Parshall Field Well Design Table 2
| Well | Lateral (Ft.) | Choke | Water (Bbls) | Proppant (Lbs.) |
| 16637 | 5293 | 20/64 | 20894 | 2183146 |
| 16461 | 4957 | 16/64 | ||
| 16543 | 5082 | 20/64 | 17796 | 2134876 |
| 16532 | 4840 | 20/64 | 19081 | 2039800 |
| 16467 | 4292 | 20/64 | 17254 | 1789276 |
| 16497 | 4472 | 20/64 | 17588 | 1738700 |
| 16635 | 4569 | 22/64 | 18197 | 1752960 |
| 16484 | 4105 | 20/64 | 17149 | 1981000 |
| 16457 | 4250 | 28/64 | 17281 | 1091000 |
| 16550 | 4433 | 20/64 | 16645 | 1919090 |
| 16671 | 4638 | 26/64 | 17250 | 1803224 |
| 16483 | 4394 | 16/64 | 17727 | 1691400 |
| 16370 | 3764 | 18/64 | ||
| 16469 | 4705 | 20/64 | 24129 | 2198291 |
| 16578 | 4753 | 20/64 | 19940 | 2052859 |
| 16371 | 4392 | 14/64 | 16791 | 1590000 |
| 16713 | 4664 | 24/64 | 18980 | 1948307 |
| 16768 | 4471 | 22/64 | 20059 | 1910795 |
| 16795 | 4308 | 20/64 | 25056 | 2624300 |
| Avg. | 4546 | 18930 | 1908766 |
Specific well production wasn't affected much by a choke difference in from 14/64 to 28/64. Water and proppant usage was important. Well number 16795 had the highest 1440-day IP rate of all the wells in this article. This well used the most water and proppant.
EOG Well Depletion: Wells With 24-Hour IP Over 1000 Bo/d
Table 3
| Well | IP Rate | 360- Day IP | 720-Day IP | 1080-Day IP | 1440-Day IP | 1800-Day IP |
| 16461 | 1487 | 469 | 410 | 321 | 264 | 231 |
| 16543 | 1015 | 440 | 321 | 261 | 226 | 197 |
| 16532 | 1285 | 489 | 365 | 291 | 253 | 223 |
| 16497 | 1675 | 502 | 364 | 292 | 246 | |
| 16484 | 1670 | 535 | 379 | 302 | 254 | 219 |
| 16550 | 1329 | 421 | 307 | 253 | 217 | |
| 16671 | 1198 | 466 | 340 | 282 | 242 | 212 |
| 16370 | 1553 | 421 | 332 | 269 | 226 | 195 |
| 16469 | 1267 | 533 | 396 | 322 | 276 | 249 |
| 16768 | 1441 | 793 | 535 | 424 | 350 | |
| 16795 | 1519 | 814 | 561 | 438 | 360 | |
| Avg. | 1404 | 535 | 392 | 314 | 265 | 232 |
Rate | 61.9% | 26.7% | 19.9% | 15.6% | 12.4% |
EOG Well Depletion: Wells With 24-Hour IP Under 1000 Bo/d
Table 4
| Well | IP Rate | 360- Day IP | 720-Day IP | 1080-Day IP | 1440-Day IP | 1800-Day IP |
| 16637 | 970 | 449 | 313 | 251 | 215 | 234 |
| 16467 | 783 | 333 | 240 | 198 | 171 | 150 |
| 16635 | 581 | 286 | 203 | 162 | 135 | 116 |
| 16457 | 922 | 481 | 339 | 275 | 232 | 202 |
| 16483 | 870 | 494 | 396 | 321 | 272 | 239 |
| 16578 | 817 | 489 | 358 | 291 | 253 | |
| 16371 | 918 | 552 | 421 | 342 | 288 | 250 |
| 16713 | 781 | 740 | 494 | 389 | 322 | |
| Avg. | 830 | 478 | 346 | 279 | 236 | 199 |
Rate | 42.4% | 27.6% | 19.4% | 15.4% | 15.7% |
In tables 3 and 4, I separated wells by 24-hour IP rates. Depletion from the first day of production to 360 days, was much higher in outperforming wells. This backs the assertion that source rock stimulation is affected more by well design, than flow from the shale's matrix. I am not saying matrix production is not affected at all, but to a much smaller extent.
EOG EURs In Parshall Field
Table 5
| Well | 1800-Day IP | Oil Produced | EUR |
| 16637 | 234 | 421200 | 842400 |
| 16461 | 231 | 415800 | 831600 |
| 16543 | 197 | 354600 | 709200 |
| 16532 | 223 | 401400 | 802800 |
| 16467 | 150 | 270000 | 540000 |
| 16497 | 203 | 365400 | 730800 |
| 16635 | 116 | 208800 | 417600 |
| 16484 | 219 | 394200 | 788400 |
| 16457 | 202 | 363300 | 726600 |
| 16550 | 179 | 322200 | 644400 |
| 16671 | 212 | 381600 | 763200 |
| 16483 | 239 | 430200 | 860400 |
| 16370 | 195 | 351000 | 702000 |
| 16469 | 249 | 448200 | 896400 |
| 16578 | 209 | 376200 | 752400 |
| 16371 | 250 | 450000 | 900000 |
| 16713 | 266 | 478800 | 957600 |
| 16768 | 289 | 520200 | 1040400 |
| 16795 | 297 | 534600 | 1069200 |
| Avg. | 209 | 376200 | 752400 |
Rate | 17.4% |
I generated the 1800-day IP rates for the wells with insufficient production (In Bold) using the average depletion. I multiplied this number by 1800 days to get the total oil produced by each well to date. The model I use estimates half of all well production is attained at 5.5 years. Being conservative, I will use five years for EURs. These numbers are very good, and derived from short laterals. If these wells were completed as 10000 foot laterals and NGLs and natural gas are figured, several would model to EURs over 2 million barrels of oil.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
Additional disclosure: Additional articles of mine can be found at shaleinsight.net. A special thanks to Craig Cooper, consultant to the oil and energy industries for his help with this article. Some of the stats and data in this article are estimates. These are only estimates and may or may not be correct depending on how these wells deplete. This is not a buy recommendation.
