You must log in to edit PetroWiki. Help with editing
Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information
Interpreting data collected during a single well chemical tracer test: Difference between revisions
No edit summary |
|||
| Line 1: | Line 1: | ||
Even with a [[ | Even with a [[Designing_single_well_chemical_tracer_test_for_residual_oil|properly designed single well chemical tracer (SWCT) test]], interpreting the data requires judgment calls, and typically, [[Simulation_of_single_well_chemical_tracer_tests|simulation]], to arrive at a final estimation of residual oil. | ||
==Overview of the area under study== | == Overview of the area under study == | ||
Tomich ''et al.''<ref name="r1" /> report one of the earliest SWCT tests, which was performed on a Frio Sandstone reservoir on the Texas Gulf Coast. The results of this test are used here to demonstrate the details of SWCT test interpretation for an ideal situation. | Tomich ''et al.''<ref name="r1">_</ref> report one of the earliest SWCT tests, which was performed on a Frio Sandstone reservoir on the Texas Gulf Coast. The results of this test are used here to demonstrate the details of SWCT test interpretation for an ideal situation. | ||
The test well in the Tomich ''et al.''<ref name="r1" /> report was in a fault block that had been depleted for several years. Because of the natural water drive and high permeability of the sand, the formation was believed to be near true ''S''<sub>''or''</sub>. When the well was returned to production (gas lift), it produced 100% water at a rate of 1,000 BWPD. | The test well in the Tomich ''et al.''<ref name="r1">_</ref> report was in a fault block that had been depleted for several years. Because of the natural water drive and high permeability of the sand, the formation was believed to be near true ''S''<sub>''or''</sub>. When the well was returned to production (gas lift), it produced 100% water at a rate of 1,000 BWPD. | ||
On the basis of observed reservoir temperature (160°F) and brine salinity [100,000 ppm total dissolved solids (TDS)], ethyl acetate was chosen as the primary tracer. Formation oil and water samples were obtained for laboratory measurement of ''K'' at reservoir temperature conditions. The value of ''K''<sub>''e''</sub> measured for ethyl acetate was 6.5 at these conditions. | On the basis of observed reservoir temperature (160°F) and brine salinity [100,000 ppm total dissolved solids (TDS)], ethyl acetate was chosen as the primary tracer. Formation oil and water samples were obtained for laboratory measurement of ''K'' at reservoir temperature conditions. The value of ''K''<sub>''e''</sub> measured for ethyl acetate was 6.5 at these conditions. | ||
==Data from the test== | == Data from the test == | ||
The test injection consisted of 1,000 bbl of formation water carrying ethyl acetate (13,000 ppm) and methyl alcohol (5,000 ppm), followed by a push bank of 1,000 bbl of formation water carrying methyl alcohol (5,000 ppm), injected over a period of two days. An eight-day shut-in period followed. During the production period, samples were collected regularly and analyzed on site using gas chromatography. The observed data are plotted as tracer concentration vs. produced volume in '''Fig. 1'''. | |||
The test injection consisted of 1,000 bbl of formation water carrying ethyl acetate (13,000 ppm) and methyl alcohol (5,000 ppm), followed by a push bank of 1,000 bbl of formation water carrying methyl alcohol (5,000 ppm), injected over a period of two days. An eight-day shut-in period followed. During the production period, samples were collected regularly and analyzed on site using gas chromatography. The observed data are plotted as tracer concentration vs. produced volume in '''Fig. 1'''. | |||
<gallery widths="300px" heights="200px"> | <gallery widths="300px" heights="200px"> | ||
| Line 16: | Line 17: | ||
</gallery> | </gallery> | ||
==Estimating residual oil saturation== | == Estimating residual oil saturation == | ||
In ideal cases, when enough data have been gathered to define the tracer profiles, it is possible to use '''Eqs. 1''' through '''3''' to approximate ''S<sub>or</sub>'' in the field. | |||
In ideal cases, when enough data have been gathered to define the tracer profiles, it is possible to use '''Eqs. 1''' through '''3''' to approximate ''S<sub>or</sub>'' in the field. | |||
[[File:Vol5 page 0623 eq 003.png]]....................(1) | [[File:Vol5 page 0623 eq 003.png|RTENOTITLE]]....................(1) | ||
[[File:Vol5 page 0623 eq 004.png]]....................(2) | [[File:Vol5 page 0623 eq 004.png|RTENOTITLE]]....................(2) | ||
[[File:Vol5 page 0624 eq 001.png]]....................(3) | [[File:Vol5 page 0624 eq 001.png|RTENOTITLE]]....................(3) | ||
If product tracer B and unreacted ester A begin together in the formation, the produced volume when A arrives back at the well (''Q''<sub>''pA''</sub>) is related to the produced volume when B arrives (''Q''<sub>''pB''</sub>) by the formula | If product tracer B and unreacted ester A begin together in the formation, the produced volume when A arrives back at the well (''Q''<sub>''pA''</sub>) is related to the produced volume when B arrives (''Q''<sub>''pB''</sub>) by the formula | ||
[[File:Vol5 page 0630 eq 001.png]]....................(4) | [[File:Vol5 page 0630 eq 001.png|RTENOTITLE]]....................(4) | ||
where ''Q''<sub>''pA''</sub> and ''Q''<sub>''pB''</sub> are in bbl. | where ''Q''<sub>''pA''</sub> and ''Q''<sub>''pB''</sub> are in bbl. | ||
This suggests that if on the same graph we plot normalized concentration of A vs. volume produced (''Q''<sub>''p''</sub>) and normalized concentration of B vs. ''Q''<sub>''p''</sub>(1 + ''β''<sub>''A''</sub>), the two curves should coincide. Because we do not know ''β''<sub>''A''</sub>, this must be done by trial and error (i.e., ''β''<sub>''A''</sub> is adjusted until the best possible match of the two profiles is found). | This suggests that if on the same graph we plot normalized concentration of A vs. volume produced (''Q''<sub>''p''</sub>) and normalized concentration of B vs. ''Q''<sub>''p''</sub>(1 + ''β''<sub>''A''</sub>), the two curves should coincide. Because we do not know ''β''<sub>''A''</sub>, this must be done by trial and error (i.e., ''β''<sub>''A''</sub> is adjusted until the best possible match of the two profiles is found). | ||
'''Fig. 2''' demonstrates this procedure. Profiles from the SWCT test ('''Fig. 1''') first were normalized by dividing each observed concentration by the peak value measured for that tracer. ''β''<sub>''A''</sub> then was varied to obtain the plot shown. The best-fit value for ''β''<sub>''A''</sub> was 0.97. | '''Fig. 2''' demonstrates this procedure. Profiles from the SWCT test ('''Fig. 1''') first were normalized by dividing each observed concentration by the peak value measured for that tracer. ''β''<sub>''A''</sub> then was varied to obtain the plot shown. The best-fit value for ''β''<sub>''A''</sub> was 0.97. | ||
<gallery widths="300px" heights="200px"> | <gallery widths="300px" heights="200px"> | ||
| Line 41: | Line 43: | ||
Using '''Eq. 3''', the ''S''<sub>''or''</sub> is approximated as | Using '''Eq. 3''', the ''S''<sub>''or''</sub> is approximated as | ||
[[File:Vol5 page 0630 eq 002.png]] | [[File:Vol5 page 0630 eq 002.png|RTENOTITLE]] | ||
Agreement is only fair over the entire curve in '''Fig. 2'''; still, this result is quite close to the final interpreted result, obtained as described below under core/simulation. The approximated ''S''<sub>''or''</sub> depends on the validity of several idealized assumptions, which are rarely satisfied in practice: | Agreement is only fair over the entire curve in '''Fig. 2'''; still, this result is quite close to the final interpreted result, obtained as described below under core/simulation. The approximated ''S''<sub>''or''</sub> depends on the validity of several idealized assumptions, which are rarely satisfied in practice: | ||
* That no B is present in the injected A, and that most of the hydrolysis reaction occurs during shut-in (no flow), so that B and unreacted A are exactly together before backflow begins. In reality, some reaction takes place during injection and production, and there always is some B present in the A as purchased. These two sources explain the high ethyl alcohol "tail" observed in '''Fig. 2'''. | *That no B is present in the injected A, and that most of the hydrolysis reaction occurs during shut-in (no flow), so that B and unreacted A are exactly together before backflow begins. In reality, some reaction takes place during injection and production, and there always is some B present in the A as purchased. These two sources explain the high ethyl alcohol "tail" observed in '''Fig. 2'''. | ||
* That ''S''<sub>''or''</sub> is uniform throughout the formation tested. In practice, however, there often is evidence of several layers with different saturations in a given completion. | *That ''S''<sub>''or''</sub> is uniform throughout the formation tested. In practice, however, there often is evidence of several layers with different saturations in a given completion. | ||
* That the injected tracer-carrying fluids are stationary in the reservoir throughout the shut-in period. Actually, fluids sometimes relocate during the shut-in period, because of pressure gradients across the field or pressure differences within subzones of the test completion. | *That the injected tracer-carrying fluids are stationary in the reservoir throughout the shut-in period. Actually, fluids sometimes relocate during the shut-in period, because of pressure gradients across the field or pressure differences within subzones of the test completion. | ||
* That the fluids inject uniformly into subzones and subsequently produce back to the test well reversibly. Total fluid flow reversibility is not observed in many tests. | *That the fluids inject uniformly into subzones and subsequently produce back to the test well reversibly. Total fluid flow reversibility is not observed in many tests. | ||
Obtaining a more reliable answer for ''S''<sub>''or''</sub> requires a detailed [[Simulation_of_single_well_chemical_tracer_tests|simulation]] of the actual test procedure for a given SWCT test. Simulators are available to account for all the nonideal situations listed above. Simulation is time-consuming and is an added expense, but in our experience, it always is justifiable and its expense is small compared to that of field data acquisition. | |||
== Nomenclature == | |||
{| | {| | ||
|- | |- | ||
|''K''<sub>'' | | ''K''<sub>''A''</sub> | ||
|= | | = | ||
| | | equilibrium partition coefficient for tracer A | ||
|- | |- | ||
|'' | | ''K''<sub>''e''</sub> | ||
|= | | = | ||
| | | oil/water partition coefficient for ester | ||
|- | |- | ||
|''Q''<sub>'' | | ''Q''<sub>''pA''</sub> | ||
|= | | = | ||
|volume required to produce tracer | | volume required to produce tracer A, bbl | ||
|- | |- | ||
|'' | | ''Q''<sub>''pB''</sub> | ||
|= | | = | ||
| | | volume required to produce tracer B, bbl | ||
|- | |- | ||
|'' | | ''S''<sub>''or''</sub> | ||
|= | | = | ||
| | | residual oil saturation, fraction of PV | ||
|- | |- | ||
|''v''<sub>'' | | ''v''<sub>''e''</sub> | ||
|= | | = | ||
|time-weighted velocity of | | time-weighted velocity of ester | ||
|- | |- | ||
|'' | | ''v''<sub>''w''</sub> | ||
|= | | = | ||
| | | time-weighted velocity of water | ||
|- | |- | ||
|''β''<sub>'' | | ''β''<sub>''A''</sub> | ||
|= | | = | ||
|retardation factor for | | retardation factor for tracer A | ||
|- | |- | ||
| ''β''<sub>''e''</sub> | |||
| = | |||
| retardation factor for ester | |||
|} | |} | ||
==References== | == References == | ||
<references> | |||
<references /> | |||
== Noteworthy papers in OnePetro == | |||
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read | Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read | ||
==External links== | == External links == | ||
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro | Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro | ||
==See also== | == See also == | ||
[[Single well chemical tracer test]] | |||
[[Single_well_chemical_tracer_test|Single well chemical tracer test]] | |||
[[Well_to_well_tracer_tests|Well to well tracer tests]] | |||
[[ | [[PEH:The_Single-Well_Chemical_Tracer_Test_-_A_Method_For_Measuring_Reservoir_Fluid_Saturations_In_Situ]] | ||
[[ | [[Category:5.6.5 Tracer test analysis]] | ||
Revision as of 12:00, 4 June 2015
Even with a properly designed single well chemical tracer (SWCT) test, interpreting the data requires judgment calls, and typically, simulation, to arrive at a final estimation of residual oil.
Overview of the area under study
Tomich et al.[1] report one of the earliest SWCT tests, which was performed on a Frio Sandstone reservoir on the Texas Gulf Coast. The results of this test are used here to demonstrate the details of SWCT test interpretation for an ideal situation.
The test well in the Tomich et al.[1] report was in a fault block that had been depleted for several years. Because of the natural water drive and high permeability of the sand, the formation was believed to be near true Sor. When the well was returned to production (gas lift), it produced 100% water at a rate of 1,000 BWPD.
On the basis of observed reservoir temperature (160°F) and brine salinity [100,000 ppm total dissolved solids (TDS)], ethyl acetate was chosen as the primary tracer. Formation oil and water samples were obtained for laboratory measurement of K at reservoir temperature conditions. The value of Ke measured for ethyl acetate was 6.5 at these conditions.
Data from the test
The test injection consisted of 1,000 bbl of formation water carrying ethyl acetate (13,000 ppm) and methyl alcohol (5,000 ppm), followed by a push bank of 1,000 bbl of formation water carrying methyl alcohol (5,000 ppm), injected over a period of two days. An eight-day shut-in period followed. During the production period, samples were collected regularly and analyzed on site using gas chromatography. The observed data are plotted as tracer concentration vs. produced volume in Fig. 1.
Fig. 1 – SWCT test results from a Frio Sandstone well.
Estimating residual oil saturation
In ideal cases, when enough data have been gathered to define the tracer profiles, it is possible to use Eqs. 1 through 3 to approximate Sor in the field.
....................(1)
....................(2)
....................(3)
If product tracer B and unreacted ester A begin together in the formation, the produced volume when A arrives back at the well (QpA) is related to the produced volume when B arrives (QpB) by the formula
....................(4)
where QpA and QpB are in bbl.
This suggests that if on the same graph we plot normalized concentration of A vs. volume produced (Qp) and normalized concentration of B vs. Qp(1 + βA), the two curves should coincide. Because we do not know βA, this must be done by trial and error (i.e., βA is adjusted until the best possible match of the two profiles is found).
Fig. 2 demonstrates this procedure. Profiles from the SWCT test (Fig. 1) first were normalized by dividing each observed concentration by the peak value measured for that tracer. βA then was varied to obtain the plot shown. The best-fit value for βA was 0.97.
Fig. 2 – Approximation of Sor using alcohol/ester superposition.
Using Eq. 3, the Sor is approximated as
Agreement is only fair over the entire curve in Fig. 2; still, this result is quite close to the final interpreted result, obtained as described below under core/simulation. The approximated Sor depends on the validity of several idealized assumptions, which are rarely satisfied in practice:
- That no B is present in the injected A, and that most of the hydrolysis reaction occurs during shut-in (no flow), so that B and unreacted A are exactly together before backflow begins. In reality, some reaction takes place during injection and production, and there always is some B present in the A as purchased. These two sources explain the high ethyl alcohol "tail" observed in Fig. 2.
- That Sor is uniform throughout the formation tested. In practice, however, there often is evidence of several layers with different saturations in a given completion.
- That the injected tracer-carrying fluids are stationary in the reservoir throughout the shut-in period. Actually, fluids sometimes relocate during the shut-in period, because of pressure gradients across the field or pressure differences within subzones of the test completion.
- That the fluids inject uniformly into subzones and subsequently produce back to the test well reversibly. Total fluid flow reversibility is not observed in many tests.
Obtaining a more reliable answer for Sor requires a detailed simulation of the actual test procedure for a given SWCT test. Simulators are available to account for all the nonideal situations listed above. Simulation is time-consuming and is an added expense, but in our experience, it always is justifiable and its expense is small compared to that of field data acquisition.
Nomenclature
| KA | = | equilibrium partition coefficient for tracer A |
| Ke | = | oil/water partition coefficient for ester |
| QpA | = | volume required to produce tracer A, bbl |
| QpB | = | volume required to produce tracer B, bbl |
| Sor | = | residual oil saturation, fraction of PV |
| ve | = | time-weighted velocity of ester |
| vw | = | time-weighted velocity of water |
| βA | = | retardation factor for tracer A |
| βe | = | retardation factor for ester |
References
Noteworthy papers in OnePetro
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read
External links
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro