CO 1

I I !

THE CARBON/OXYGEN LOG MEASUREMENT

ABSTRACT

BY P. A. WICHMANN, E. C. HOPKINSON,

and V. C. McWHIRTER

Dresser Atlas - Houston, Texas

A new pulsed neutron hgging technique has been de- veloped to unambiguously identify the presence of the element carbon. The CarbonlOxygen Log is currently available on a limited commercial basis in North America and has been

i

L i I

I tested under a wide variety of field conditions. f

This paper covers the theory of the logs response and shows examples of what is available in the way of taped output. The basics of interpretation by a simplified pattern recognition technique are reviewed along with test pit measurements and varified field response.

INTRODUCTION

During the past three years, a Carbonloxygen logging

survey several hundred wells. Despite the fact that a great 1 (;$ many of these applications have been made as a last resort 1 when definitive answers could not be derived from other more I conventional logging methods, the device has scored a very

high success ratio in predicting the type of fluid ultimately produced. Also complicating these initial efforts is the fact that development and improvement of the instrumentation has been continuous over this period of time and even now, a significant advance is on the verge of commercial introduction.

I device has been commercially available and has been used to

I

y Rays From Inelastic Collisions

1000 11s

Jf & Days cI y Rays From Thermal Neutron Capture y Rays From Neutron Activation Products

FIGURE 1

The Dresser Atlas Carbon/Oxygen Log is made possible by an electro-mechanical device that can be pulsed in extreme- ly short, concentrated bursts of 14 Mev neutrons and a detector system that can selectively discriminate and count the high energy gamma rays that result from the inelastic collisions of these neutrons with the atoms of carbon and oxygen. dj Figure 1 schematically reviews the source and time frame of neutron bombardment induced gamma radiation. Note that the inelastic or prompt gamma rays exist only while the very

high energy neutron source is on and for a very few micro- seconds following the cessation of the pulse. From this period of time until the neutron source is reactivated, only gamma rays of thermal neutron capture, and to a lesser degree, activation gamma rays, are present. This is an extremely important point, as the 4.43 Mev inelastic carbon gamma ray is the only unambiguous nuclear indicator of that element in nature.

THE INSTRUMENT AND ITS CALIBRATION

The current commercial version of the C/O logging instrument is made up principally of the neutron generator, an electro-mechanical apparatus using the deuterium on tritium reaction to produce 14 Mev neutrons at an average rate of 108 per second and a gated NaI scintillation detector. The window is 6 to 7 microseconds wide and the repetition rate is 4,000 neutron pulses per second. The current generation of instru- ments is designed for stationary readings to overcome the statistical limitations of the data.

A prototype device, designed after the method of Schultz et al, is currently being field tested as a continuous logger. To overcome the current statistical limitations of the older instru- ments, the generator is pulsed more rapidly at 20,000 times per second and the resultant increased undesired background is compensated for by measuring just prior to the neutron burst. Figure 2 is a short section of a 35% porosity sand containing both oil and water that was logged at 4 feet per minute with this prototype instrument.

FIGURE 2

THE LOG ANALYST 25

The commercial stationary C/O instruments provide all the background of experience on which our interpretation data base is predicated. The same phenomena that produces the unique 4.43 Mev inelastic carbon gamma ray also produces a 6.1 Mev inelastic oxygen gamma ray. The surface calibrator is comprised of a conventional chemical neutron source inside its carrying container. This container is made of wax with an outer iron casing. This assemblage of material in proximity to a conventional neutron source insures the presence of 7.6 Mev iron, the 4.43 carbon and 2.22 Mev hydrogen gamma rays. Figure 3 shows the typical spectrum of a surface calibration of this device. The most prominent peaks of gamma ray energy apparent from this illustration are the inelastic photo peak of carbon and its two escape peaks at 4.43 MeV, 3.92 Mev and 3.41 MeV respectively, and the hydrogen capture peak at 2.2 MeV. Also identifiable in this calibration spectrum are the iron capture peak and its two escape peaks at 7.64 MeV, 7.13 Mev and 6.62 MeV, respectively.

Hydrogen Capture 2.2 MEV . CH. 64

20 40 60 80 100 120 140 160 180 200 220 240

f Scale Change

FIGURE 3

Other gamma rays are present but not particularly prominent in this surface calibration. The typical locations of each of these gamma ray peaks on the 256 channel analyzer are summarized in Table I.

TABLE I

Identifying Gamma Ray Energy-Mev Channel Number

Silicon Inelastic Calcium capture Hydrogen capture Carbon 2nd escape Carbon 1 s t escape Carbon photo peak Iron 2nd escape Iron 1 s t escape Iron capture peak

1.78 1.96 2.2 3.41 3.92 4.43 6.62 7.13 7.64

5 1 57 64 99

113 128 189 204 218

This surface calibration has always been easily maintained downhole as we have consistently been able to identify at least three elements on a continuous basis as the measurements are being made. Iron is always present in both the instrument housing and the casing, the hydrogen capture peak is always prominent and either silicon or calcium, depending on the

major rock type present, have also been consistently in evidence.

In field operation the two detector gates are chosen to encompass the more prominent escape peaks of both the 4.43 Mev carbon and the 6.13 Mev oxygen inelastic reactions.

The reason for this is apparent in Figure 3, which shows the lower energy escape gamma rays to be more numerous than are the gamma rays in the original photo peaks. In operation, the total number of counts in each of these two gates are cumulated with the carbon total divided by the oxygen total. This is illustrated with the typical inelastic spectrum shown in Figure 4. The shaded areas represent the carbon and oxygen gates. The taped data may also be dis- played in the tabular form illustrated in Table 11.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

20 40 60 80 100 120 140 160 180 200 220 240

Channel Number

FIGURE 4

TEST RESULTS

These carbon-oxygen (C/O) ratio numbers have proven to be quite significant in defining the fluid content of known rock types, regardless of the salinity of the connate water in the system.

Test pit and test barrel results illustrate the stability, repeatability, and consistency of the numbers. First, 55-gallon drums containing loosely packed clean beach sand-porosity 40%-were fitted with a 4" O.D. pipe through their axial centers. Each of three such sand-filled drums were saturated with kerosene, alcohol (C,H,OH) and fresh tap water. The C/O ratios measured in these three barrels are summarized in Table 111.

Subsequent field measurements of clean or shaly (but non-carbonaceous) 100% water bearing sands have consistently recorded C/O values of 1.57 to 1.6 1 . As the percentage of bulk volume of oil increases, the C/O ratio has consistently in- creased in approximately the same proportion that one would predict from these simple sand barrel tests illustrated in Table 111.

26 NOVEMBER-DECEMBER, 1976

TABLE II DATA FROM CRRBOWOXYGEN LOG RUN 7

CHANNEL NO.

, 0 10 28 30 45 50 60 70

38 180 I10 I20 I30 148 158 I60 1 PI3 180 190 206 210 220 236 240 258

80

1

0. 9.

8082. 3634. 2385. 1865. 1183. ae4. 841. 628. 513. 446. 365. 360. 322. 316. 353. 225.

137. 91. 73. 46. 33. 23. 15.

1136.

2

8 .

7335. 3360. 2277. 1767. 1121. 913. 783. 700. 532.

354. 345. 292. 345. 335. 265. 201. 142. 67.

42. ? I . 18. 14.

80 .

4013.

se.

3

0. 367.

6633.

2069. 1615. 1123. 913. 717. 589. 510. 424. 351. 327. 343. 383. 277. 274. 197. 130.

55. 42. 23. 16. 16.

32e7.

138 .

4

0. 7664. 2817. 3194.

1433. 1166.

697. 625. 543 * 451. 343. 386. 357. 274. 297. 280. 160. 133. 79. 47.

2s . 18. 1 1 .

2888.

869.

38.

5 6

a. 0. 28915. 23101.

f d f l . 5408. 3141. 3015. 1929. 1858.

1141. 1039.

-. -

1410. 13137.

913. 1322. 682. 701. 509. 589. 536. 525. 486. 451. 368. 371. 340. 326. 345. 355. 288. 290. 267. 2613. 276. 252. 143. 154. 124. 105. 71. 63. 45. 54. 32. 40. 34. 18. 13. 22. 20.

7

8. 13947. 4738. 2768. 1797. 1343. 1019.

667. 552. 519. 419. 367. 326. 347. 273. 244. 247.

98. 6 3 . 43. 31. 7 2 . 26.

830.

1713.

8

1 . 11348. 4483. 2660. 1766.

993. 825. 659.

470. 376. 342. 322. 321. 254. 282. 215. 163. 107. 57. 42. 28. 31. 19.

12132.

585.

9

2. 10059. 4265.

1798. 1301. 978.

663. 556. 472. 397. 316. 274. 318. 391. 256. 195. 143. 95. 56. 37. 39. 28 . 16.

25113.

1306.

I 0

5. 9 1 1 1 . 3878. 2435. 1777. 1266. 920. 794. 655. 521. 487. 375. 354. 319. 334. 310. 256. 181. 132. 117. 52. 54. 30. 2n. 12.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RUll NUMBER DEPTHCFT) CAL CHI EN 1 CRL CH2 EN2 TIME(SEC1

7 0.0a 64.00 2.22 206.08 7.13 360

UINDOW I = 3.27- 4.24 MEV WINDOW E = 4.98- 5.137 MEV

LOWER CHANNEL UPPER CHANNEL SUM ST DEV LOWER CHANNEL UPPER CHRNNEL SUN ST DEV RRTIO ERROR TOTRL COUNTS

94. 122. 13996. 118.30 141. 169. 8928. 94.49 1.5613 8.021 204871. 94.37 122.42 13914.13 117.96 141.51 169.56 890e.09 94.38 1.562 0.021 204071.

TABLE 111

s a a !-

0 0 .

Fluid C/O Ratio

Kerosine barrel Alcohol barrel Water barrel

1.86 1.70 1.61

1.72

1.70

1.68

1.66

1.64

1.62

1.60

1.58

1.56 100 80 60 40 20

Water Saturation %

FIGURE 5

A typical saturation calibration curve-from a California sand sequence with an average porosity of 30%-illustrates this simple linear relationship in Figure 5.

When non-carbonaceous sands are the rule, interpretation is just that simple. High C/O values indicate a large bulk volume of oil and low values show water. Table IV summarizes some C/O data from a non-carbonaceous heavy oil bearing sand and compares it to a core analysis of the same interval. Each logged station was repeated with the C/O measurement and multiple core analysis measurements were also made.

Core Analysis Depth 785

7 59

740

723

702

683

647

c/o 1.61 1.61 1.60 1.61

1.63 1.64

1.66 1.67

1.56 1.57 1.68 1.68

1.69 1.68

SiICa $J 1.65 31.3

34.5 1.67 32.4

32.3 31.5

1.66 29.5 27.2 30.1

1.69 31.4 32.8 31.7

1.44 Shale

1.64 32.5 31.5 31.2

1.60 31.1 30.2 30.7

TABLE IV

Oil Water 0.0 78.4 0.0 81..7 26.9 70.0 23.8 76.4 16.4 82.0 22.2 71.2 21.5 71.5 34.3 55.2 42.5 57.1 42.3 55.1 43.9 53.5

54.6 41.0 53.8 45.1 47.4 46.2 60.2 39.0 57.6 41.9 60.2 38.7

THE LOG ANALYST 27

Besides the increasing C/O ratio with increasing oil saturation, please pay special attention to the C/O number in the shale at 702 feet. The close agreement of this value to those in 100% water bearing non-carbonaceous sands is the reason that the measurement is so useful in shaly sands as well as those containing fresh to brackish formation water. Non- carbonaceous shales have been consistent in exhibiting responses on this order of magnitude.

ADDITIONAL LOG DATA

The C/O measurement uses data that is recorded only while the neutron source is active. For the station measure- ment instrument, this represents only a little more than 3% of the total time. Between pulses, a large number of gamma rays of thermal capture are present. These can be very useful in defining key elements, notably silicon and calcium, that type the actual rock matrix. To take advantage of this capture phenomena the other half of the 5 12 channel analyzer is used to accumulate gamma rays between the neutron pulses.

with this data and either the total capture gamma ray counts or the inelastic gamma ray counts may also be recorded. These records have also been useful in helping to define the presence of gas.

FIGURE 7

SIMPLIFIED FIELD INTERPRETATION

Where carbonates and tighter limey sands occur in proximity to productive non-carbonaceous sands, both C/O and Si/Ca data must be used to define the productive intervals. Lock and Hoyer proposed a simple pattern recognition tech- nique to accomplish this interpretation. Figure 8 represents their plot of C/O and Si/Ca data from an East Texas well. The circular symbols represent known gas sands in the area. Gas, with its lower concentration of carbon per unit volume of hydrocarbon material, yields values of the C/O ratio between those of oil and water bearing rocks. Based on test data and known production, Figure 9 was then sectioned into areas of likely production characteristics that are quite accurate for this specific area. As the lithology becomes more complex and water salinities change (gamma rays of capture for chlorine occur at coincident peaks with the calcium) the interpretation becomes more complex and computer techniques and addi- tional logs that are described elsewhere must be used.

4.0

3.0

$ 2.0 fa E E

c?i 1.0

0 2 4 6 0

Gamma Ray Energy . MeV

FIGURE 6

1.8

Figure 6 represents idealized capture gamma ray spectra for both a limestone and a sandstone. Note that the major silicon peaks (the dotted curve) occur in roughly the same part of the spectrum as do the inelastic gamma rays resulting from carbon. Conversely, the calcium peaks are most prevalent in the area where the inelastic oxygen gamma rays are measured. In a similar manner to the C/O ratio measurement, a Si/Ca ratio is also taken as a stationary reading. Because of the larger number of counts available, this measurement may also be recorded continuously as a lithology indicator. Figure 7 shows such a continuous log of the silicon and calcium curves and their ratio correlated to an electrolog. Also included on the log is the trace of the source monitor. Our procedures are flexible

0 1.7 F a a 0 0 1.6 .

1.5

1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0

Si/Ca RATIO

FIGURE 8

28 NOVEMBER-DECEMBER, 1976

1.8

1.7 0 I-

2 0 o 1.6 .

1.5

8. Vexcellha, W. C., Skaggs, L. L, Bishop, W. D., DeVries, M. R, McWhirter, V. C, and Wichmann, P. A, Clastic Formation Evaluation Program in Cased Holes for Both Fresh and Saline Formation Waters to Determine Hydrocarbon Content., SPE Fall Meeting, Dallas, 1975, SPE 5508.

McWhirter, V. C, Introduction to Carbon Logging, SPE Re- gional, Casper, 1976, SPE 5906.

9.

1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1

SVCa RATIO

FIGURE 9

CONCLUSIONS WICHMANN

The C/O log is an important new technical development in the search for hydrocarbons that are already cased off. It also provides for a direct measurement of the key element carbon, making the salinity less critical parameter in cased lt of this instrumentation has been cununuuus m u LIK currently available stationary measurement will soon be giving way to tools that will be capable of continuous logging.

of the connate water amuch hole log analysis. Developmer -_- ---L. -__-__- *LA --_---

ACKNOWLEDGMENT

The authors must acknowledge the work of a number of Dresser Atlas employees, but most particularly that of R. B. Culver and R. A. Bergan, who were most responsible for the instrumentation and neutron source development that are the heart of th is logging system.

REFERENCES

1. Hoyer, W. A., and Rumble, R C., Field Experience in Measuring Oil Content, Lithology and Porosity with a High Energy Neutron- Induced Spectral Logging System, Journal Petroleum Tech- nology, July, 1975, p. 801.

2. Caldwell, R. L, Mills, W. R., Jr., and Hichman, J. B., Jr., %amma Radiation From Inelastic Scattering of 14-Mev Neutrons by the Common Earth Elements, Nuclear Science and Engineering (1960), Vol. 8, p. 173.

3. Lawson, B. I,., and Cook, C. F., and Owen, J. D., A Theoretical and Laboratory Evaluation, presented at the 45th Annual Meet- ing of the SPE of AIME, Houston, Texas (October 47,1970).

4. Lock, G. A. and Hoyer, W. A. and Interpretation, Journal 19741, p. 1044.

, Carbon-Oxygen (C/O) Log: Use of Petroleum Technology, (Sept.

5. Schultz, W. E. and Smith, H. -., -vvr..rvry liul,. Evaluation of a CarbonlOxygen (C/O) Well Logging System, Journal of Petroleum Technology, (Oct. 1975), Vol. XXVI.

6. Smith, H. W. and Schultz, W. E., Field Experience in Deter- mining Oil Saturations from Continuous C/O and Ca/Si Logs Independent of Salinity and Shaliness, Transactions of SPWLA Fifteenth Annual Logging Symposium, June 2-5, 1974, McAllen, Texas.

7. culver, K. B., Hoplunson, P;. C. and youmans, A. H., --carbon/ Oxygen (C/O) Logging Instrumentation, SPE Journal, VoL 14, No. 5, Oct. 1974.

THE LOG ANALYST 29

McWHl RTER HOPKINSON

ABOUT THE AUTHORS

M R P. A. WICHMANN is the Manager of Log Analysis for Dresser Atlas in Houston. A 1958 honors graduate Petroleum Engineer from the Colorado School of Mines, he worked for the Shell Oil Company from that date until 1965 in a variety of positions, including Petro- physical Engineer. At that time he went to work for Dresser as a Research Log Analyst for Lane-Wells. In 1968 he became Chief Log Analyst for Dresser Atlas and assumed his present duties early in 1975.

Mr. Wichmann has been active in the SPWLA for a number of years, and was recently on the National Board of Directors. He has authored or coauthored over 30 technical papers, including 16 that have appeared in SPWLA publications. He has also spoken at numerous international and local SPWLA and SPE meetings. Besides the SFWLA, he is a member of the CWLS, the SPE of AIME, Tau Beta Pi, and is a Registered Engineer in the State of Texas.

MR. VERNIE C. McWHIRTER is Senior Projects Engineer in the Research and Engineering Department of Dresser Atlas. He has been associated with Dresser Atlas and its predecessors since 1951.

He has had extensive field experience in nuclear logging techniques and has written several papers on nuclear logging techniques.

He attended Texas Tech University in Lubbock, Texas.

M R ERIC C. HOPKINSON is Manager of Pulsed Neutron Logging Systems in the Research and Engineering Department of Dresser Atlas.

He has been associated with nuclear logging research since 1952 when he joined Well Surveys, Inc. which later became a part of Dresser Industries. Mr. Hopkinson has had an important part in the develop- ment of a number of Dresser Atlas nuclear innovations, including the Neutron Lifetime Log and the C/O Log.

He graduated from Queens University in Kingston, Ontario, Canada with a BSc in Engineering Physics in 1950. From that time until he joined Well Surveys, he worked for the Canadian Government Department of National Defense.