Introduction

Well testing has been used for decades to determine essential formation properties and to assess wellbore condition. There are many different types of tests that can be utilized to collect this information depending on when the test is conducted, the well location, the well type, and the formation type. For the most part, conventional tests (flow/buildup or injection/falloff) have satisfied the majority of our needs. However, under certain conditions, traditional test methods are not feasible for various reasons. This is especially true for very low permeability formations that require massive stimulation to obtain economic production. For these formations, it is extremely important to establish the formation pressure and permeability prior to the main stimulation. One test that has proved to be convenient for this purpose is commonly referred to as a "minifrac" test.

Overview

A minifrac test is an injection-falloff diagnostic test performed without proppant before a main fracture stimulation treatment. The intent is to break down the formation to create a short fracture during the injection period, and then to observe closure of the fracture system during the ensuing falloff period. Historically, these tests were performed immediately prior to the main fracture treatment to obtain design parameters (i.e. fracture closure pressure, fracture gradient, fluid leak-off coefficient, fluid efficiency, formation permeability and reservoir pressure). However, since personnel and frac equipment were all waiting on location to perform the main treatment, the falloff period was usually stopped shortly after observing closure, before reliable estimates of formation pressure and permeability could be obtained. Since these two parameters are critical to the fracture design and for production/reservoir engineering, it seemed prudent to extend the falloff period to obtain better estimates, especially since there is little hope of gathering this information after the main stimulation. Many operators have accomplished this by simply scheduling the minifrac test well ahead of the main fracture treatment.

Predicting the falloff time required to obtain meaningful estimates of formation pressure and permeability is difficult, as it depends on having prior knowledge of the permeability, in addition to knowing the geomechanical properties of the formation. In normal and/or over pressured reservoirs that will support a full column of water in the wellbore, the progress of a minifrac test can be assessed with pressure data measured at the wellhead, eliminating the need for "guessing" when sufficient data has been obtained. In under pressured reservoirs that cannot support a full liquid column in the wellbore, the test progress can only be monitored with wellhead data until the fluid level starts to fall in the wellbore. In these situations, bottom hole pressure measurements must be obtained, and longer shut-ins are generally required to overcome wellbore dynamics resulting from the falling liquid level.

Analysis Techniques

There has been much research conducted by numerous experts to establish or advance techniques for analysing data obtained from minifrac tests. Due to the nature of a minifrac test, the analysis is conducted in two parts; Pre-Closure Analysis (PCA) and After-Closure Analysis (ACA). Similar to traditional Pressure Transient Analysis (PTA), specialized time and derivative functions are utilized to perform PCA and ACA. Variations of these analysis techniques are used in commercial fracture simulation software, which is not convenient for everyday use.

The most common minifrac analysis techniques used in the industry today have been implemented in IHS WellTest. When combined with the efficient data management, dynamic wizards and superb graphical user interface within IHS WellTest, users have the ability to easily analyse minifrac test data. An additional benefit gained with IHS WellTest, is the ability to advance after-closure analysis beyond diagnostics and straight line analysis, to include after-closure modelling capability. This provides the advantage of verifying and improving on results obtained from diagnostic analyses, similar to the analysis workflow used for many years in traditional PTA.

Unlike traditional PTA, which is founded on the "constant-rate solution", the main ACA techniques are founded on the "impulse solution". However, both of these solutions provide similar results if they are correctly applied. In IHS WellTest, two different ACA techniques have been implemented, which have their roots in both solutions. These techniques are based on the work of K. Nolte and M. Soliman. Nolte's is a common technique implemented in fracture simulation software, and has unique diagnostic capabilities. Soliman's technique is more consistent with traditional PTA methods, and facilitates the use of analytical models (based on D. Craig's work) to simulate after-closure pressure behaviour. The main references for these solutions can be found within the IHS WellTest help topics. Both ACA techniques yield similar estimates of pressure and permeability if a consistent analysis is performed.

Observations from Real Data

Observations made from minifrac tests conducted in various shale/tight formations in North America have provided valuable insight on what to expect. Not surprising, the time required to achieve radial flow during the falloff period is indicated to be greatly influenced by the injection rate/volume. Higher injection rates/volumes create farther reaching fractures, and because a significant portion of the fracture usually remains open after the main closure event, linear flow dominates for a much longer period. Moreover, high injection rates/volumes increase the chance of creating multiple fractures, and depending on the proximity to overlying/underlying pay zones, increases the chance of contacting multiple zones. Hence, minimizing injection rates/volumes can simplify both the pre/post closure analysis, thus improving on estimates of reservoir pressure and formation permeability.

In many cases, injection (typically water) rates on the order of 1 to 2 bpm (1440 - 2880 bbl/d) have been observed to quickly achieve formation breakdown and propagate a fracture, with average injection times of 2 - 5 minutes (after breakdown). Fracture closure is normally observed within 1 to 24 hours of falloff, which is mainly controlled by the magnitude of "effective" formation permeability. In higher permeability formations, radial flow can develop quickly after closure, while in lower permeability linear flow can dominate for long periods.

When radial flow develops quickly, a falloff period on the order of 1 - 3 days is sufficient to obtain reliable estimates of reservoir pressure and formation permeability. However, when linear flow is observed after the main closure event, the development of radial flow is delayed, and may not be observed in a reasonable time-frame. Although this introduces greater error in estimating reservoir pressure and formation permeability, extending the falloff period will help reduce the error. In these situations, both permeability and fracture extension strongly influence the required falloff duration. Lower permeability and longer fracture lengths extend the time required to achieve radial flow. Even if radial flow is not achieved, the "upper and lower limits" of reservoir pressure and formation permeability can usually be determined from the after-closure model, which is very useful.

An example of a minifrac test analysed using IHS WellTest is shown in figures 1 - 7. The test was conducted on a vertical well at a formation depth of 10,000 ft. Pressures were monitored at the wellhead, and converted to bottomhole values for analysis. The total test duration is about 24 hours.

Figure 1 shows the pressure profile during the 18-minute injection period performed at 1 bpm (1440 bbl/d). The sudden drop in pressure (shortly after injection commenced) indicates that formation breakdown occurs very quickly. Although the injection period of this test is significantly longer than normally prescribed for an unconventional play, the formation permeability is high enough (in this case) to yield reasonable results.

Figure 2 shows the pre-closure analysis using the semilog and first derivative corresponding to G-function time. From this plot, fracture closure is identified within the initial 3 -hours of the falloff period.

Figure 3 represents the log-log diagnostic plot used to confirm fracture closure and identify the flow regimes developed "after-closure". In this case, the semilog derivative (calculated with respect to shut-in time) exhibits a slope of -1 shortly after closure, suggesting that radial flow has developed. The fluctuations in the derivative slope can be attributed to disturbances from pockets of gas/air moving slowly up the wellbore, which cause small changes in the measured wellhead pressure. These disturbances can be avoided or minimized by ensuring the wellbore is completely liquid filled, and that air is purged from the surface equipment prior to the pump-in period.

Figure 4 represents the falloff data plotted using Nolte's radial time function, and Figure 5 represents the falloff data plotted using Soliman's radial time function. When radial flow is achieved after closure, both of these plots should yield a "constant slope". However, as indicated from Figure 3, the radial flow trend is disturbed; causing the data to "shift". However, a constant slope is maintained, and by fitting a straight line through the radial flow trend developed between these disturbances (Δt ≈ 15.57 hr), both analyses yield the same extrapolated pressure of 5333.4 psia, and kh on the order of 0.1 md.ft (k = 0.002 md using the formation thickness h = 60').

Figure 6 depicts the log-log plot of the derivative corresponding to Soliman's solution. The solid line represents the simulated derivative obtained with the after-closure model. The model suggests radial flow was not quite achieved during the test period, and would likely develop after about 50 hours of shut-in (dashed line displayed after the test data points). This contradicts the diagnostic analysis, which suggests radial flow had been achieved, and demonstrates the value of coupling traditional diagnostics with an after-closure model.

Figure 7 depicts Soliman's radial time plot, including the simulated pressures (solid line) corresponding to the after-closure model shown in figure 6. Since the model suggests radial flow is not achieved, different estimates of kh and Pi are obtained, and in this case are higher (0.18 md.ft and 5440 psia, respectively) than those calculated from the diagnostic analysis. It should be noted that in cases when linear flow dominates the after-closure pressure behaviour, the after-closure model will yield lower estimates of the kh and Pi.

Martin Santo, Senior Technical Advisor, IHS Energy
Posted December 3, 2015

This article was published by S&P Global Commodity Insights and not by S&P Global Ratings, which is a separately managed division of S&P Global.