Infill program revitalizes Wyoming gas field

Feb. 11, 2002
A diverse group with different technical backgrounds and from different organizations successfully solved problems that had plagued past attempts to revitalize sour-gas production from the Carter Creek field in Wyoming.

Based on a presentation to the SPE Rocky Mountain Petroleum Technical Conference, Denver, May 21-23, 2001.

A diverse group with different technical backgrounds and from different organizations successfully solved problems that had plagued past attempts to revitalize sour-gas production from the Carter Creek field in Wyoming.

In the early 1990s, an infill drilling program was unsuccessful due to high costs and poor production performance of the new wells. But late in 1997, a joint Chevron USA Production Co. and Halliburton Energy Services team put together a program that reduced new well costs by more than 35% and obtained initial producing rates from new wells that rivaled those of wells completed in 1982.

Geology

Whitney Canyon-Carter Creek, 15 miles north of Evanston, Wyo., was the first field to produce hydrocarbons from the Paleozoic section of the Wyoming Thrust Belt. The field was discovered in 1977 during an intensified exploration effort that followed earlier Thrust Belt discoveries.

Click here to enlarge image

The five productive formations in the field consist of the Ordovician Bighorn (dolomite), the Mississippian Lodgepole (dolomite), the Mississippian Mission Canyon (dolomite), the Pennsylvanian Weber (eolian sandstone), and Triassic Thaynes (limestone). The Mission Canyon, the primary reservoir in the field, contributes nearly 75% of the reserves.

Hydrocarbons in Whitney Canyon-Carter Creek reservoirs are structurally trapped by anticlinal closures within the hanging wall of the Absaroka thrust plate. A back-limb imbricate, commonly known as the Bridger-Hill thrust, divides up the sections of the Absaroka fault to create the field's eastern margin.

Hydrocarbons sourced from Cretaceous shales beneath the Absaroka thrust are sweet; however, all formations except the Thaynes produce sour gas due to a chemical-thermal hydrocarbon reaction with the abundant anhydrite throughout the Paleozoic section.

Initial wells

Spudded in 1976, the Amoco-Chevron-Gulf Working Interest No. 1 (ACG WI No. 1) discovered the Whitney Canyon/Carter Creek field. The well tested 910 Mcfd of gas on a drillstem test (DST) from the Triassic Thaynes limestone. After drill pipe parted while coring the Permian Phosphoria, the operator completed ACG WI No. 1 the Thaynes formation in August 1977, and tested it at 4.7 MMcfd of gas with 88 b/d of condensate.

Click here to enlarge image

In December 1978, the operator completed the ACG WI No. 2 in the Bighorn dolomite, flowing 5.7 MMcfd of gas and 64 b/d of condensate. The well also tested the Mississippian Mission Canyon at 8.5 MMcfd of gas and 130 b/d of condensate.

Intense field development at Whitney Canyon-Carter Creek occurred between 1980 and 1983, driven by developing the field on a competitive basis. Through mid-2001, the field had 68 wells and sidetracks in or around producing structures.

The two most productive wells in the field, both on the Whitney Canyon structure, were:

  1. Amoco-Champlin 457-A1, which had initial potential production of 65 MMcfd of gas in 1979.
  2. Chevron Federal 1-6E, which produced 55 MMcfd of gas after being recompleted in the Mission Canyon in 1989. It previously produced from the Bighorn.

The Mission Canyon reservoir in the Whitney Canyon structure has good matrix permeability (1 to 5 md) and an open natural fracture system, which contributes greatly to flow. The Mission Canyon reservoir in the Carter Creek structure is also fractured, but calcite and anhydrite heal most of the fractures. As a result, the highest rate wells in Carter Creek are slightly above 20 MMcfd of gas.

Click here to enlarge image

Peak gas production in the 1980s for Whitney Canyon-Carter Creek field was about 320 MMcfd. In January 2001, field production was about 240 MMcfd.

Opportunity

The Carter Creek gas plant has been operated mostly at full capacity, but production decline in recent years created excess capacity and provided an opportunity to add new wells to accelerate reservoir depletion and add incremental reserves.

During 1998, the joint technical team evaluated opportunities and made and implemented recommendations for a program to revitalize the field. The Chevron team members included a production engineer, a reservoir engineer, a drilling engineer, a facilities engineer, a geologist, and a geophysicist.

Halliburton, for its part, placed a technical advisor and a drilling engineer in Chevron's Evanston office and provided support from its Denver offices by allocating a reservoir engineer, a geologist, and a geophysicist to work part time on the project.

Click here to enlarge image

The team meets every other month in the Evanston or Denver offices and has developed a good working relationship.

Reserves

The Mission Canyon reservoir is a normally pressured, undersaturated, sour-gas-condensate reservoir under depletion drive. The produced gas contains 15% H2S with 5% CO2. The initial reservoir pressure measured was 6,100 psi at 14,500 ft with a condensate yield of 22 bbl/MMcf.

Current reservoir pressure is about 2,700 psi with a condensate yield of 10 bbl/MMcf. More than 550 bcf of gas have been produced from Carter Creek structure.

The technical team applied material-balance calculations, well performance, and volumetric analysis to determine the recoverable remaining reserve potential. It then compared these various estimates with one another to determine the reserve potential for new infill wells.

In general, well performance (production decline-curve analysis) showed fewer remaining reserves than material-balance calculations, thus indicating the need for additional infill drilling. In addition, the volumetric estimate was larger than the material-balance number. Again, this suggested the need for development drilling.

The technical team then selected new well locations by two different methods.

The first method identified offsetting wells, which have the largest remaining recoverable reserves and the longest remaining well life. Offsets to these wells would demonstrate current completion effectiveness when compared to the originally completed wells.

Click here to enlarge image

The second method targeted offsetting poorer producing wells (under performers) where rock properties are similar to average wells. Offsets to these wells would add reserves in addition to showing the benefit of current stimulation practices.

In general, in both cases, the new infill wells (with half the drawdown pressure) initially produced at similar or higher rates than the originally completed wells. Hence, the stimulation effectiveness today is twice as good.

Completions

The initial well completions in the Carter Creek Mission Canyon reservoir, 1980-1983, combined small acid treatments with larger acid fracturing jobs. Both techniques are common for completing wells in carbonate reservoirs. These jobs effectively removed damage and created some reservoir stimulation. Nodal analysis indicated that these treatments resulted in skin values of 0 to 21.

A 1990 development program used a similar completion approach and obtained similar skin values. But, because 13 years of production had reduced reservoir pressure to 4,000 psi from the initial 6,100 psi, producing rates were lower than for the field's initial wells.

Click here to enlarge image

The rates were insufficient to justify additional infill wells, and therefore, the operator did not continue the program.

In 1997, the operator launched a second infill well program. The initial completion strategy aimed at improving zonal coverage by using mechanical isolation to break the stimulation job into multiple smaller stimulations. The jobs consisted of perforating a small reservoir interval, acid fracturing, isolating the treated zone with a composite plug, and drilling out the plugs once the entire productive zone was treated.

Click here to enlarge image

At the beginning of this program, the reservoir pressure was 2,700 psi or about a 3.5 ppg equivalent mud weight. The partially depleted reservoir condition was a major concern for the completion team.

To deal with low pressure, the program called for pumping acid as a CO2 foam. The foam energized the fluid for cleanup and also assisted with diversion.1 The program also called for removing the viscous pad in order to minimize water in the system. In addition, the operation used a snubbing unit for drilling out the isolation plugs, and most treatments achieved a pump in rate of 40 bbl/min.

Although promising, the results were still poor. The primary problems were that:

  • Multiple treatments and plugs increased job complexity and cost
  • External isolation of the acid was difficult since the reactive fluid tended to seek out the lower pressure, higher permeability zones

To deal with these problems, the operator formed a technical study team to evaluate and develop a better understanding and techniques for stimulating the Mission Canyon.

Study results

Mission Canyon's main porosity zone (MPZ) is a low permeability (1.5-0.1 md) dolomite. Based on a brief study of core data from several offset wells, the team divided the MPZ zone into the following four units (Fig. 1):

  • Upper MPZ A, which has primarily moldic porosity and is very tight.
  • MPZ A, which has good porosity (9%), medium permeability (0.7-0.5 md), and is highly fractured. Unfortunately, most of its fractures have been healed, primarily by anhydrite with some calcite (Fig. 2a).
  • MPZ B, which produces about 65% of the gas that comes from the Mission Canyon. Its porosity is similar to the MPZ A (10%) but it has slightly higher (1.5-1.0 md) permeability. The unit is highly fractured but the fractures tend to be filled with calcite (Fig. 2b).
  • Lower MPZ B, which is similar to the MPZ B in porosity and apparent permeability. Its fractures, however, tend to be filled with anhydrite, similar to the MPZ A.

    The production log (Fig. 1) shows the relative flow contribution from each zone.

    After the core study, the team reviewed pumping results from recently completed wells and observed that hydraulic fracturing was not physically occurring. The formation would not crack (frac fluid could not be pump into the well) after perforation of the well in acid and application of a maximum wellhead treating pressure (6,500 psi) with water in the hole.

    Click here to enlarge image

    During an acid treatment, however, once the reactive HCl acid entered the perforations, the pump rates slowly increased to as much as 50 bbl/min. At this rate, it had been assumed that the formation was parting. Such behavior is not characteristic of fracturing, however, because once one exceeds a formation's fracture gradient, the formation should take fluid regardless of fluid type.

    Because of the slow-rate-increase phenomenon, the team questioned if the reservoir was being fractured. The question remained on how such high rates (40-50 bbl/min) were possible in a low reactive, low permeability (1.0 md) dolomite with only a matrix acid job. A review of the cores led to the answer.

    Click here to enlarge image

    The original core study tried to correlate open fractures with conventional logs, but the team discarded the need for the correlation upon determining that many fractures in the MPZ were healed. After the team combined core data with the pump-in data, it became apparent to them that the critical factor was the fill in the fractures and not the fractures themselves.

    The team concluded that HCl acid reacting with calcite fill, such as in MPZ B, resulted in the high pump in rates.2 In such sections, acid pumped at below-fracture pressure would attack the most reactive rock, namely the calcite in fractures (Fig. 2c), and form wormholes in the calcite. Because the lower reactive dolomite trapped the calcite on both sides, the team hypothesized that wormholing would proceed in two dimensions along the old fracture plane (Figs. 3a-3c).

    At greater than 32° F., calcite dissolution is mass transfer limited.3 The team, therefore, recommended high pump rates in an attempt to place live acid deeper into the formation. Very high rates have been shown to cause more excessive wormhole branching in calcite, consuming more acid for a given penetration depth. The team did not see excessive branching as a problem because cleanout of the entire fracture left more dolomite matrix open to flow.

    Click here to enlarge image

    The Carter Creek dolomite dissolution is reaction-rate limited because of low reservoir temperature (205° F. at 14,500 ft).3 In this case, the team did not want to dissolve dolomite, only the calcite in the fractures; therefore, it selected a retarded acid system to slow dolomite dissolution.

    The retarded acid combined with high pumping rates lowers the Damkohler number, thus, impeding spending against the dolomite walls (reaction-rate limited) and increasing wormholing penetration in the calcite (mass-transfer limited).

    Because the dolomite is tight (1.0-1.5 md), the team did not worry about losing significant amounts of unspent acid to the dolomite matrix, which could create a spongy rock structure and waste acid.3

    Lab work could prove out this hypothesis. The team, however, was in the middle of a development program and had to put the model into practice by making the following changes to the treatments:

    • Doubling pumping rate to 80 bbl/min minimum.
    • Using gel acid for retardation.
    • Completing a well as one single large interval.
    • Increasing perforation density and depth.
    • Using floating ball sealers for diversion by dropping the balls spread out, not in clumps.
    • Using large, early acid stages (MPZ B) and smaller acid stages later in the job.
    • Using foam acid with 50% CO2.
    • Bringing a well back on production as soon as possible.

    Well 3-31M, completed with the new program, averaged 22.5 MMcfd during the first month. Fig. 4 compares the production of the best original development well, Chevron Federal 1-29M, with Well 3-31M. Note that the Carter Creek gas plant causes the 1-29M's production spikes.

    Nodal matches for the two cases (Figs. 5a and 5b), show a 0 skin for 1-29M and a -4 skin for 3-31M.

    The operator drilled three additional wells (Chevron Federal 2-32M, 2-29M, and 2-5F) because of the program's success. The operator leveraged what was learned from the first three wells and improved the flow rates from the second set of wells (Figs. 6 and 7).

    The average production rate of the new wells is 15 MMcfd. The team also focused on cost control (Fig. 8). The change to one single large stimulation led to numerous cost efficiencies. For instance Well 3-31M's completion cost was $1.55 million down from $4.4 million on the first well.

    Drilling

    Wells drilled in the overthrust belt of Wyoming are deep and the formations hard and abrasive. Wells in the late 1970s and early 1980s often took more than 1 year to drill. The first well in the new infill program was 17,096 ft deep and took 144 days to drill from spud to rig release. The team, however, evaluated new ways to further reduce drilling time.

    Various proposed penetration rate equations have similarities.4 Penetration rate is proportional to weight on bit (WOB) and bit rpm and is inversely proportional to mud weight.

    Hydraulics and tooth wear also play important roles while formation characteristics are something that cannot be changed.

    Because the intermediate hole has a salt section, most operators have drilled that section with salt saturated mud. Salt saturated mud has considerably higher mud weight than necessary, therefore, slowing penetration rates.

    A change to oil-based mud for drilling the well would lower mud weight, thus improving rate-of-penetration (ROP) and reducing washouts in the salt. Fewer washouts would reduce cement costs. Also a change to straight-hole mud motors would increase the bit rpm, thus increasing penetration rate.

    But a concern remained about the effect of greater rpm's on insert and bearing wear. The team also considered increasing WOB but decided against it due to bearing wear concerns and mud motor limitations.

    Federal 2-32M was the first well drilled with oil-based mud and some straight-hole mud motors. These changes considerably decreased the days needed to drill the well. The team fully implemented the program on Federal 2-29M, resulting in a decrease of almost 30 drilling days, on a comparable depth basis (Fig. 9). Cost reduction amounted to $1.3 million even with the additional costs for oil-based mud and straight-hole mud motors.

    The team also evaluated the possibility of improving bit selection. The formations unfortunately are too hard for polycrystalline diamond compact (PDC) bits.

    Detailed dull bit grading correlating the bits to the formation help selecte the bits that improved performance.

    At times, drilling required a new bit design. Once the bits were optimized, the team discovered that when formation changed, it was more cost effective to trip and run the proper bit in the well than to drill, for what could be several days, at a slower than optimum penetration rate.

    The production-liner's cement job was one problem in completing the wells. In the wells, several producing zones have pressures that vary by several thousand psi. Tracer surveys on the first completions showed that the acid tended to move to the low-pressure zone rather than stay in the perforated zone.

    The team decided to use a foam cement system to reduce these problems, although it did have some concern that the cement would not hold up to the stimulation pressures. A detailed evaluation of the cement bond logs and tracer surveys showed that the foam cement is an excellent solution.5

    Facilities

    Site and facilities costs had been between $1.2 and $1.4 million, depending on terrain, and required earthwork for large rigs, well depth, and the reclamation after completion.

    In the past, operators installed aboveground the field's gathering system and required electric heat trace and insulation for the wells.

    In 1990, operators started burying lines but these still required insulation and vapor barrier.

    Site costs remained the same for the 1997 well, the Chevron Federal 2-6F. Much of site costs relate to the fact that around the field there is no usable base material. All base and finish material for the site and access roads must be trucked over long distances.

    The team redesigned the facilities for this well to a new standard by using some surplus equipment. The redesign used buried gathering lines and eliminated one building.

    Although these changes reduced costs, facilities and site costs still remained about $1.2 million. To cut costs, the team evaluated skid-mounted equipment and making new wells satellites of existing facilities. The 3-31M facility was the first to include a standard satellite design, piggy backing the use of a test system and flare from an existing facility.

    Facility construction and flowline installation included lump-sum contracts for a new design that encompassed buried flowlines 8 ft deep (below frost line), eliminating the insulation, and using fusion-bonded epoxy coating.

    These steps helped reduce facility and site installation costs to $959 million. This satellite concept on the next two wells, 2-32M and 2-29M; further cut costs.

    Using different material for the oil pit was another change on these two wells. The material increased cost more than $8/ton but had less water content than the previous material. This change lowered cost by about $30,000 because less material was required for solidification.

    Also, previous operations had sent the water from the pits to the Carter Creek plant for disposal. Solids entrained in the water caused the plant considerable expense in cleanup because of settlement in the ponds.

    The alternative of hauling this material to a disposal site would have cost $10/bbl.

    Aeration using surplus equipment located at the plant brought down the disposal costs to less than $1/bbl.

    The changes made have reduced site and facilities costs for the new wells to an average $650,000, or about $550,000 less for each new infill well).

    Total costs for each well are now about $5.2 million compared to the previous $9.9 million. The economic impact of these cost savings and the increased production rates have been dramatic (Fig. 10).

    The initial well had a 14% rate of return (ROR) whereas the last well had a 146% ROR.

    Acknowledgments

    The authors would like to thank Chevron's operations personnel and to Kevin Kopp, Scott Russell, Sheila Sullivan, and Ron Lackey for their contribution to this article. The also are also grateful to the management of Chevron USA Production Co. and Halliburton Energy Services Inc. for allowing the publication of this article.

    References

    1. Halliburton, Carbonate 2020 Best Practices.
    2. Fredrickson, S.E., "Stimulating Carbonate Formations Using a Closed Fracture Acidizing Technique," Paper No. SPE 14654, 1986.
    3. Hoefner, M.L., and Fogler, H.S., "Reaction Rate vs. Transport Limited Dissolution During Carbonate Acidizing, Application of Network Model," Paper No. SPE 15573, SPE Annual Technical Conference and Exhibition, New Orleans, October 1986.
    4. Burgoyne, Millheim, Chenevert, and Young, Applied Drilling Engineering, SPE, Dallas.
    5. Kopp, et al., "Foamed Cement vs. Conventional Cement for Zonal Isolation - Case Histories" Paper No. SPE 62895, SPE Annual Technical Conference and Exhibition, Dallas, October 2000.

    The authors

    Click here to enlarge image

    Brett Buzarde is a production engineer for Chevron USA Production Co. in Evanston, Wyo. Buzarde has a MS in petroleum engineering from Texas A&M and a BS in petroleum engineering from the University of Tulsa.

    Click here to enlarge image

    Gary L. Guymon is a facilities representative for Chevron USA Production Co. in Evanston Wyo. He has been a project or construction manager for various Chevron companies. Guymon has a master electrical license.

    Click here to enlarge image

    Craig Seal is a principal reservoir engineer working in integrated solutions for Halliburton Energy Services Co. in Denver. Seal has a BS in petroleum engineering from Colorado School of Mines. He is a member of SPE and is a registered professional engineer in Texas.

    Click here to enlarge image

    Kelly Olds is a lead engineer for Halliburton Energy Services in Denver. Olds has a BS in chemical engineering from the University of Utah. He is an SPE member and is a registered professional engineer in Texas and Utah.

    Click here to enlarge image

    Jay M. Foreman is a principle technical professional for Halliburton Energy Services in Denver. Foreman has a BS in mechanical engineering from the University of Missouri-Rolla.