Deciding on which wells are best candidates for having their performance improved is not always easy.
As pointed out in a recent Gas Technology Institute workshop on restimulation technology, one key problem is that many operators make restimulation decisions without benefit of downhole data and base their decisions only on production information, which may not be a good indicator.
In other words, because of such constraints as manpower and capital, the industry tends to restimulate the worst wells and not those that have the most potential, thus leading to a negative perception of restimulation.
GTI study
The GTI workshop primarily centered on its 21/2 year study on hydraulically refracturing tight gas wells.
It found that refracs are only 2-3% of total US fracs, and most of these refracs involve gas wells. The US Midcontinent, Rocky Mountains, and South Texas regions have more than 85% of these refraced gas wells.
The GTI study focused on developing new, efficient candidate-selection methodology and short-term, low-cost verification tests for refracturing tight gas wells. The work included laboratory studies to investigate potential damage mechanisms and advanced restimulation concepts and demonstration tests in three regionally dispersed tight gas fields.
It estimates that 15% of the wells in a field represents 85% of the restimulation potential.
After it analyzed 200-300 wells in each area, GTI refraced four wells in the Frontier formation of the Green River basin of Wyoming, two wells in the Mesaverde formation in the Piceance basin of Colorado, and three wells in the Cotton Valley formation in the East Texas basin.
Although all demonstration refracs were not successful, GTI said the program did determine that refracs in the right wells could succeed if the work follows proper procedures, such as including an appropriate fracturing fluid, adding nitrogen or carbon dioxide gas to assist in clean-up, and using perforation ball-outs for limited entry.
Its candidate selection included individual well screening for mechanical well integrity, pressure analysis, economics, et al., and wells previously damaged by ineffective gel breakers or no proppant clean-up during initial fracs.
Methodologies
In its evaluation, GTI separated well performance into reservoir and completion components and picked candidate wells based on two or three of the following methodologies: 1.) Production statistics obtainable from public data, 2.) virtual intelligence with detailed well data, and 3.) type curves, which require more data and considerably more interpretation.
Each method yielded different candidate wells, and GTI therefore did a benchmark study on a hypothetical field to gain insight into each method's effectiveness for ranking refrac candidates. One conclusion was that none of these methods accurately estimated reservoir properties or incremental reserves on a quantitative basis.
Regarding their effectiveness, GTI found:
- Production statistics identified underperforming wells relative to offsets but overlooked highly productive wells that can be superior candidates. It found a good correlation between the best 12 months of production and ultimate recovery, although benchmark analysis indicated the method was ineffective for selecting refrac candidates.
- Virtual intelligence provided insights into performance drivers but lacked analytic rigor. The method involved pattern recognition in a three-step process of building a neural network, creating a generic algorithm, and applying fuzzy logic. GTI's benchmark analysis indicated that, although the method has potential, it needs more development to be reliable.
- Type curves identified highly productive wells with restimulation potential but oversimplified well conditions, and the method was labor-intensive. For identifying high-potential restimulation candidates in complex, multilayered tight gas reservoirs, benchmark analysis found this method to be the most reliable.
The use of early-time shut-in data was one option discussed at the workshop to lessen the cost and time for obtaining type-curve data.
This method brackets the permeability and fracture length solution but does not give an exact answer. Well shut-in periods that include the start of linear flow is a recommended minimum criterion for the method.
Other work
Various operators, notably Patina Oil & Gas Corp., Denver, have successfully refraced the Codell gas-condensate reservoir in the Wattenburg field area in Colorado. Patina selects refrac candidates with an algorithm that considers a weighted average of such factors as formation porosity-feet, gas-oil ratio, peak production, cumulative production, expected ultimate recovery, and differences in the ultimate recovery from offset wells.
Patina's estimates that 1,100 wells have been refraced in the area and that 4,575 potential refrac candidates remain.
Its 1999 program refraced 110 wells, at $98,500/well, resulting in 30,500 boe/well being added to gross reserves. Patina showed, at the GTI workshop, that after starting to include a diagnostic algorithm, refrac results improved markedly.
