|
The Cook Inlet gas play from the perspective of a small independent working in Alaska
Andy Clifford Special for Petroleum News This section is available to oil and gas companies to tout their North America prospects. Word count: 500-2,000 words with 1-2 graphics (map, photo, chart, etc.) Email submissions to [email protected] or call 907 770-3505. In a previous article (see Oct. 17 issue of Petroleum News) Aurora Gas shared its thoughts regarding the Lower Tyonek and Hemlock oil play. Aurora is just beginning to mature its exciting inventory of onshore oil prospects with drilling expected to begin in 2005. In this article, Aurora is focusing on the shallow gas play, which it has successfully exploited in the last four years with established gas production in four separate fields, Kaloa, Lone Creek, Moquawkie and Nicolai Creek, all on the west side of Cook Inlet.
The Cook Inlet sedimentary basin of Alaska covers an area of approximately 12,000 square miles. A cumulative total of almost 1.3 billion barrels of oil and 7 trillion cubic feet of gas have been produced in the basin since the first discovery was made at Swanson River in 1957. Most of the exploratory drilling in the basin was undertaken prior to 1968, when larger reserves were discovered on the North Slope of Alaska and industry’s attention and technology focused to the north. Nearly 90 percent of the present reserve base was discovered in that early phase of drilling.
Thermogenic gas in-place for Cook Inlet has been estimated at 3.1 tcf. A further 7.5 tcf of biogenic gas reserves have been estimated to be present. Cook Inlet non-associated gas within the Sterling, Beluga and Tyonek formations is a mixture of biogenic “swamp” methane and gas from low grade bituminous coal degasification formed and trapped in situ in nearby channel sands. Such gas is formed independent of the deeper thermogenic gas associated with the oil fields. The gas gravity for all formations is approximately 0.56.
The Tyonek gas sands are generally more productive than the Beluga sands but less than those of the Sterling formation. The majority of the Sterling gas fields have exceptionally good rock quality with the formation’s thick, blocky sands. Aurora’s gas fields produce gas from Beluga and Tyonek reservoirs. Aurora believes there isn’t enough exploration There is no shortage of gas in Cook Inlet; or rather there should not be a shortage of gas. In Aurora’s opinion, there is plenty of gas yet to be discovered in the basin but there are not enough exploration dollars being spent, nor enough companies exploring the basin’s potential. There are several reasons for this:
There is a perception in industry that the remaining gas potential in Cook Inlet is limited due to a lack of appreciable exploration success in recent times.
Good quality seismic data is scarce to non-existent over large tracts of the basin and is generally limited to 2D seismic, much of which was acquired prior to the 1990s.
The geology is complex both structurally and stratigraphically. The steeply dipping flanks of the wrench-induced anticlinal folds are poorly imaged on seismic data and the prospective non-marine sands have limited areal extent and are also poorly imaged because of the numerous interbedded coals that “mask” the sands.
Log analysis in Cook Inlet is particularly difficult since commonly used techniques such as neutron-density crossover only work for thicker, cleaner sands such as those in the Sterling and parts of the Tyonek formations. Most gas pay in Cook Inlet has to be interpreted using a combination of log curves. Many of the old wells in the basin only have a sonic log for porosity estimation. In addition, reserve estimation is very difficult, especially without the benefit of multiple wells and good pressure data. Costs of Cook Inlet exploration high The costs of exploration in Cook Inlet are very high compared to areas in the Lower 48. Most seismic acquisition requires the use of heliportable drilling, which is expensive, and operations on the west side of the inlet require the use of barges, planes, landing zones, staging camps and portable satellite communication towers. Development of shallow, low-pressure gas reservoirs such as those at Aurora’s Kaloa, Lone Creek, Moquawkie and Nicolai Creek fields, where productive depths range from 1,300-3,500 feet, requires the use of compression and dehydration facilities on the surface and gravel packs and sand screens in the subsurface to prevent sand production from the loose, unconsolidated reservoirs. Such procedures are expensive and mechanical surface equipment is subject to occasional failures.
The discontinuous non-marine sandstones require more development wells, which increase the cost of development. New discoveries of gas will require gas gathering lines and pipelines to connect with existing infrastructure.
Access to gas infrastructure and markets is restricted which is a major deterrent to new players.
So, given all of the above, why would somebody like Aurora Gas be interested in exploring for and developing gas in Cook Inlet? Aurora believes remaining gas potential ‘outstanding’ Aurora believes the remaining potential for discoveries of non-associated biogenic gas in Cook Inlet is outstanding. There have been some excellent recent successes by Marathon and Unocal at Ninilchik and Happy Valley showing there are fields of 50-100 billion cubic feet still to be found in Cook Inlet. It is worth noting, however, that both of these successes were spurred by pre-existing wells at Falls Creek and Happy Valley with tested gas or bypassed gas pay recognized on logs. The early successes were then expanded into other areas on trend.
Aurora has identified a multitude of bypassed gas opportunities throughout the basin in old wells drilled for deeper oil objectives as well as new play types without productive analogs in the basin. Two specific examples of these are Aurora’s Three Mile Creek and Long Lake prospects. The former prospect, where Aurora is currently drilling the Three Mile Creek Unit-1 well, is targeting Beluga and Tyonek gas reservoirs updip from the Superior Three Mile Creek-1 well, drilled in 1967. The structure is a four-way anticline located in the footwall of the Moquawkie Anticline, which lies to the west. Aurora recently completed pre-stack depth imaging of newly acquired 2D seismic data over the prospect. The Long Lake area is one where Aurora is pursuing Tyonek-aged alluvial fan deposits with gas trapped against basement granodiorites that form the Cottonwood-Stedatna Creek pluton. The Tyonek sands are much thicker with excellent reservoir characteristics in this area and the potential for large gas reserves is very exciting. Aurora re-evaluated the 1973 Texaco Long Lake Unit-1 well with a re-entry of the well and new logging during 2004 and plans a follow up well in early 2005. Existing seismic expensive to license Existing seismic data is expensive to license (typically $1,000-2,500 per mile), invariably of poor quality and difficult to license in some areas because of confusion surrounding ownership of the data plus a lack of incentive for some companies to license data. New acquisition costs can run as high as $25,000 per mile for 2-D and $100,000 per square mile for 3-D. Aurora recently helped usher in the use of mulchers for seismic acquisition, a technological application commonly used in Canada and parts of the Lower 48 for “minimal impact” seismic, with a cost reduction of about 10-15 percent.
The cost of seismic processing has dropped dramatically of late and Aurora has taken advantage of this by doing reprocessing of existing seismic, both 2-D and 3-D. One of their 3-D surveys has undergone several reiterations in the last 12-18 months. Illumination studies, ray-tracing and modeling can help determine the necessary line length and offset needed during seismic acquisition. The extra cost of adding more shot holes to lines to adequately image the deep structure has been a deterrent to companies. Furthermore, because the objective of older seismic acquisition was often deeper oil targets, there was very little common depth point fold in the shallower gas-prone section. Getting adequate high fold in the latter requires a denser grid of shot holes, which are expensive. Complex geology includes structural inversion The complex geology can be better understood by studying analogs from other basins and from within Cook Inlet itself. The west side of Cook Inlet shows strong evidence of structural inversion. Ongoing studies by Aurora at Moquawkie and Nicolai Creek using recently acquired 3-D seismic data suggest that the existing anticlines were once structural lows, as evidenced by isopach mapping of discrete intervals. These structural lows acted as a focusing mechanism for anastomozing fluvial channels prior to uplift and inversion in the Plio-Pleistocene.
One of the first things Aurora ever did when entering the Cook Inlet gas play was to study existing fields such as Kenai and Beluga River and in particular the log responses within those fields to gas sands and wet sands. They made a catalog of log responses for each formation and have subsequently updated it. Aurora has also undertaken substantial log analysis of gas sands and wet sands from existing fields as calibration wells for their prospects. They have also compiled an inventory of seismic characteristics of existing gas pay, including seismic “gas chimney” effects and direct hydrocarbon indicators.
Aurora acquired two 3-D surveys over Moquawkie and Nicolai Creek fields in 2003 and has experimented with a revolutionary amplitude vs. offset, AVO, technique that helps differentiate coals (which have a unique AVO signature), gas sands and wet sands. Just as with log interpretation, where the first step is to recognize the coal signature, the coals can be “blackened” out, leaving a seismic dataset with just gas sands, wet sands and shales, each having unique seismic characteristics. Then by time-slicing at various flattened seismic horizons that correspond to gas pay or wet sands in wells within the dataset, Aurora can start to map individual sand body geometries such as fluvial channels within the Beluga and Tyonek formations, thereby potentially overcoming one of the problems in Cook Inlet, not being able to differentiate stratigraphy. Aurora aims to test its hypothesis in 2005 by targeting stacked AVO anomalies in a stratigraphic trap setting in what might be a first for Cook Inlet exploration! Log analysis presents challenges Gas sands are typically detected by using neutron-density logs. Whereas for clean, normally pressured sands, the presence of gas is evident wherever crossover of the neutron porosity and density porosity logs occurs, this is not always the case in Cook Inlet. Log analysis in Cook Inlet is particularly challenging, especially with respect to the gas play. While the crossover method utilizing neutron and density logs usually works quite well in clean formations, or those without heavy mineral particles, it provides very pessimistic results for deposits such as the Beluga formation.
In general, the natural presence of gas will cause a shift in the density porosity readings to higher values while the neutron porosity readings will be lower than the density readings. A commonly used technique in Cook Inlet is to make a shift of 12 porosity units when overlaying the neutron and density logs. Another commonly used technique is Schlumberger’s Elemental Log Analysis, “ELAN,” which uses a software model designed specifically for Cook Inlet. It provides readings for the amount of bound water in the formation, the amount of gas present and amount of moved gas.
Aurora makes extensive use of log analysis by NuTech Energy, who run conventional digital log data through a three-step process for editing, normalizing and correction of any calibration errors then model the response to provide new irreducible water saturation or BVI outputs and finally produce a simulated nuclear magnetic resonance or NMR log. The multi-track output they provide includes permeability, volumetrics, lithology, clay volume displays and two flag tracks.
NuTech has helped Aurora identify bypassed gas pay in Cook Inlet. The crossover method works in cleaner, thicker sands such as the Sterling formation. Very rarely, neutron-density crossover is seen in Beluga sands, but they are usually too thin or too shaley. Tyonek sands can exhibit crossover in thicker, cleaner sands but otherwise resemble the Beluga interval.
Cycle skipping on the sonic log is also generally a good indicator of gas pay, especially in the cleaner Sterling and Tyonek sands. Formation resistivity for proven gas pay varies greatly throughout Cook Inlet but deep resistivity should generally read more than 20 ohms, although there are notable examples, especially within the Tyonek where resistivity is as low as 13-14 ohms for wells that have produced more than 15 bcf of gas. However, in such rare cases there is usually a good mud log show even if neutron-density crossover is lacking. Recognition of gas-water contact important Whilst log analysis poses particular challenges in identifying gas pay to begin with, further evaluation of a gas reserve is difficult without recognition of a gas-water contact. Such contacts are readily seen in Sterling sands and thicker Tyonek intervals, such as the Grayling gas sands of McArthur River field. Volumetric analysis for the Beluga and Tyonek formations is extremely difficult since location of gas-water contacts are hard to delineate. In many cases there might not be a true, distinct gas-water contact at all but rather a large transition zone between water saturations of 30-40 percent in the gas leg and 60-70 percent in the water leg.
There appears to be a lot of “fizz” gas or low saturation gas, probably caused by continuing gas recharge from the interbedded coal beds.
By far the most challenging problem facing log analysts and petrophysicists working Cook Inlet is correct calculation of water resistivity. Variations in formation water resistivities are believed to be the main cause of water saturation cutoff variation between wells. Water resistivities calculated from the SP, spontaneous potential, curve are possibly not valid because of SP reductions related to shaliness effects and hydrocarbons.
Other problems affecting volumetric reserve estimation include: (1) the lack of stratigraphic definition on seismic data; (2) the presence of seismic gas chimneys; (3) low seismic fold at shallow depths; and (4) poor seismic resolution at depth due to energy absorption by the coals and generation of interbed multiples by the coals.
Gas reserve estimation is best accomplished by material balance, supplemented by decline analysis, where pressure data is sparse, and finally volumetrics. Extended well tests with about 300 million cubic feet of gas production are usually sufficient to give a good estimation of recoverable reserves. Many wells in Cook Inlet have commingled completions making it difficult to allocate production from existing fields to individual sands within any given field. Furthermore, net pay thicknesses and reservoir parameters vary greatly for gas reservoirs in Cook Inlet. High exploration and operating costs due to low activity levels There is very little that can be done to reduce exploration costs in Cook Inlet.
Because there is such little activity with few players, there is a limited number of contractors, a limited number of rigs, no offshore rigs and a limited supply of technology (both logging, completion and seismic). There is also a restricted operating season for some activities.
Aurora wanted to acquire multi-component 3-D seismic to use shear data to better see through the “gas clouds” over some of their fields but the equipment was prohibitively expensive to bring in. Operation costs can be reduced by the concept of areawide development with initial wells bearing the brunt of expensive gas gathering pipelines and facilities with subsequent wells being more economic because of short tie ins and shared operational support. Multi-well drilling/completion programs can benefit from the bundling of contractor services.
Aurora has developed gas reservoirs shallower than anyone else in Cook Inlet. Unfortunately, such reservoirs are low pressured and require sand control as well as compression in order to meet pipeline specifications. Recovery averages 85 bcf per well for Sterling sands The discontinuous areal extent of such Beluga and Tyonek sands means that more wells are required for development.
The Sterling sands are the most prolific in Cook Inlet with producing depths ranging from 3,300 feet-5,200 feet. Recoverable reserves range from 1.9 bcf per well at Beaver Creek to 216 bcf per well at North Cook Inlet, with an average of 85 bcf per well. The field decline rate for Sterling fields ranges from 7-20 percent with an average of 10 percent. The Beluga sands have porosities ranging from 10-22 percent and producing depths of less than 1,300 feet to 8,100 feet.
Aurora has established the shallowest Beluga production yet in Cook Inlet at its Nicolai Creek Unit No. 9 well, where the shallowest production is from less than 1,300 feet. Recoverable reserves range from 4.5 bcf per well at Lewis River to 18 bcf per well at Kenai with an average of 13 bcf per well. The decline rate for Beluga fields ranges from 14-36 percent with an average of 17 percent.
The Tyonek reservoirs generally exhibit porosities of 19-22 percent with producing depths of 1,700 feet to 9,000 feet. Recoverable reserves range from 0.5 bcf per well at Granite Point to 90 bcf per well at McArthur River with an average of 56 bcf per well. The average decline rate for Tyonek fields ranges from 14-25 percent.
The shallower gas reservoirs are generally unconsolidated making coring somewhat difficult and most conventional and sidewall coring to date has been focused on the deeper oil and gas reservoirs. Water saturations for proven gas accumulations range from 35-50 percent with recovery efficiencies reaching as high as 90 percent.
Gas reserves per well depends on a number of factors including the lateral extent of sands, contributions from coal beds, the number of productive sands but also on depth and pressure. Test rates for Tyonek sands generally average 5-10 million cubic feet per day with some higher initial rates in some McArthur River wells.
Recoveries for the respective formations range from 960-1240 thousand cubic feet per acre-foot for Sterling sands, 600-1,150 mcf per acre-foot for Beluga sands and 490-1,150 mcf per acre-foot for Tyonek sands.
To date, Aurora has constructed five gas gathering pipelines. One of these, connecting the three wells at the southern end of Nicolai Creek to the Cook Inlet Gas Gathering System, required the tunneling of a section from the top of the bluff near the Granite Point Tank Farm under the adjacent wetlands to the beach. The Lone Creek and Moquawkie fields have longer pipelines of five to six miles each, which will substantially enhance the economics of future developments in the vicinity of each field. Future development prospects will also be able to benefit from sharing of surface facilities. Gas marketing presents challenges Gas is a very different commodity from oil and the gas market in Southcentral Alaska is a “closed” market without access to Lower 48 markets. Several large fields such as McArthur River (Grayling Gas Sands), Kenai, Beluga River and North Cook Inlet, have dominated Cook Inlet gas production for decades. Producers and utilities alike have “entrenched” positions and it is difficult for new producers in the basin to gain access to key infrastructure since the pipelines do not have open access as “common carriers.”
In addition, there is considerable uncertainty in the gas market with the possibility of future closures of both the Agrium fertilizer plant and the liquefied natural gas facility, thereby changing the whole market dynamic.
Aurora Gas recently brought its fourth gas field on stream with the successful tie-in of the Kaloa-2 well. Exploration drilling has begun at the Three Mile Creek Prospect and Aurora plans a full year of drilling activity for 2005 and is excited by the under-explored gas potential in the basin.
Editor’s note: Andy Clifford is vice president of exploration for Aurora Gas LLC.
| 