Comparison of DRC's THP - BHP Models Vs. A Commercially Available Model

By Chad Cluver | Wed, 7 May 2008

When converting surface pressure measurements to bottomhole pressure, the capability of the software model is of utmost importance in assuring accurate and repeatable conversion.  Which factors the model takes into account and how the inputs affect the model can be the difference between getting data that is useful for analysis and data that is inconclusive.  Data Retrieval Corporation has developed the most rigorous conversion model in the industry.  This model can account for varying amounts of liquids produced, varying tubing size, a wide range of gas compositions, and a wide range of reservoir temperatures.  The model can even account for a phenomenon known as wellbore thermal decay which is explained below.  The question is how does this model for wellhead pressure to bottomhole pressure conversion stand up to other commercially available software?

Figure 1 illustrates the challenges associated with wellhead pressure data when considering converting to bottomhole pressure data.  The well depicted in the chart is high rate (greater than 50,000 MCFD), has a bottom hole temperature of about 200 deg F, and produces some condensate (about 30 BBL/MMCF) and essentially no water.  It can be seen that the wellhead data is subject to many factors that can influence it in such a way as to make PTA analysis seem impossible.  A model’s ability to account for these factors when calculating BHP is crucial to getting an accurate conversion that can be used to obtain useful reservoir characteristics.  

Figure 1:  Wellhead vs Bottomhole Pressure  

Figure 2 compares the calculated BHP’s of DRC’s model with those of a commercially available software package.  As can be seen in the chart, the commercially available software has a difficult time matching the measured flowing bottom-hole pressures (black curve) as recorded by a downhole gauge, especially at increasingly higher rates.  It also couldn't match the shut-in BHP's.  The main cause of this error is incorrect thermal modeling of the wellbore, which can be broken down into two aspects.  The first is thermal decay, which is the cooling of the wellbore fluids upon shut-in resulting in increasing density of the fluid and thus a decrease in pressure at the wellhead.  The second is wellbore warming, which occurs when the rate is increased and the wellbore temperature increases.  Both of these effects can easily been seen in Figure 2.  Even when using a wellhead temperature profile which takes into account thermal decay and wellbore warming, the commercially available package is unable to accurately calculate BHP.  The build-up curve (green and purple curve) is in fact distorted due to the incorrect thermal modeling.

Figure 2:  BHP Comparison Plot  

Incorrect thermal modeling leading to an incorrect conversion to BHP will lead to incorrect PTA analysis.  A build-up that doesn't properly account for thermal decay is distorted and an incorrect analysis would be difficult to detect.  It is impossible to know how a model matches actual bottomhole data without prior experience with which to base the conversion on.  DRC has conducted hundreds of tests per year for the past 22 years and has an extensive pool of past experience that ensures accurate conversion of wellhead pressure to bottomhole pressure.  Without similar experience, an operator would be basing his conversion on faith that the software will be correct, not a proven history of accurate conversion.  It is important to note that the importance of accurate thermal modeling is more pronounced as permeability increases.

Another thing that is important to consider is the speed of the conversion from WHP data files to BHP.  The commercially available software being compared is significantly slower in doing this, and actually has a limitation of 32,000 data points that can be converted at one time.  Our in-house software has no such limitations; we often convert data in excess of 100,000 points.  Using the Black Oil PVT method took about 25 minutes to convert 32,000 points, and using the EOS SRK PVT method was SIGNIFICANTLY slower, which is why in Figure 2 that data set is incomplete.  It took almost 16 hours to convert just 3,800 points using EOS which extrapolates out to about 5 1/2 days for all 32,000 points.  As a contrast, our in-house software can run this conversion in a matter of seconds.

While it is true that there are commercially available software packages on the market today that do offer the ability to convert WHP to BHP, accuracy and time constraints make them a less desirable option over what the DRC SPIDR Surface Well Testing System offers.  DRC conducts several hundred well tests each year and has a 22 year history to back up the accuracy of the conversions.  Our conversion algorithm is the most accurate over the largest range of conditions available today.  When this is combined with our ability to turn around a conversion and analysis the same day that the data is received it is easy to see why so many companies are turning to DRC for their well testing needs.