 |
|
|
Newsletter Sign-Up
Testimonials
The SPIDR device is attractive to us at EOR for many reasons. For our St. John’s CO2/Helium project, the information we gather in the evaluation phase of the reservoir is valuable for future development plans going ...
- Jim Skurner, P.E. 
|
General News
General Data ProcessingBy Rod Kelly | Thu, 9 Jun 2005 The purpose of this article is to provide an overview of the different types of models/correlations that DRC uses in order to provide both accurate BHP's and accurate analysis. The topics covered will be:
The SPIDR is only used to acquire data. Once we have obtained the wellhead pressures, we convert them to bottomhole conditions. Most engineers who deal with natural gas are aware of the Cullendar/Smith or Gray correlations for calculating bottomhole pressure from surface pressures. At DRC, we have modified and expanded the validity of these models considerably and can now test almost any gas or gas/condensate well that naturally unloads its produced liquids. Once we have completed the bottomhole conversion, we then perform elementary analysis. FrameworkThe backbone of all of our gas and gas/condensate BHP algorithms is the Cullendar-Smith model. The Cullendar-Smith correlation is accurate for single phase, dry gas. It assumes an average temperature and average z-factor, along with laminar flow to perform its calculations. DRC has adjusted the Cullendar-Smith correaltion by accounting for produced liquids, segmenting the tubing (with each element having its own fluid properties), and by allowing for multiple flow regimes (laminar, intermediate, turbulent, highly turbulent flow). The result is a much more rigorous (and accurate) treatment of the wellbore physics, which continues to be our single-phase gas/gas-condensate model.Multi-phase flow correlationsWhen the pressures in the wellbore are below the dew point for a gas-condensate system, there will be some form of 2-phase (or 3-phase, if water is present) flow. If the well produces with enough gas rate (velocity) to continually unload itself naturally (avoid slugging), then the pressure drop in the wellbore/tubing can be calculated using either our Mist Flow or Annular Mist Flow models. We have also developed a Bubble Flow model for oil wells, though at this point we are still in the R&D phase with this model. Wellbore Thermal and Phase Behavior TransientsWhen any well undergoes a buildup or a drawdown, the well bore is subjected to a thermal transient. If the well has been shut in for an extended period of time, the temperature profile in the well bore closely resembles the surrounding geothermal gradient. When that well is brought on line, fluids enter the well bore at reservoir temperature and lose heat through the tubing wall on the way to the surface. Eventually, the well bore will assume a new thermal profile based on the flow rate of the well and the geothermal flux. The reverse is true on shut-in in which case the well bore cools.If Pressure Transient Analysis (PTA) is to be performed using surface measurements, the ability to model these changes in well bore temperature is essential. DRC has worked with several Major E&P companies in acquiring thermal data from high BHT wells with which to develop this Thermal Decay model. DRC has over 15 years experience testing wells in which ThermalDecay was a necessary consideration for correct analysis. The WHP to BHP conversion routines which incorporate thermal decay cannot be used on surface data files that have been acquired by instruments that are subjet to ambient thermal response. It is impossible to de-couple thermal decay effects from thermal compensation effects in a a gauge which is responding to both at the same time. The failure to compensate for these effects on moderate to high permeabiilty wells results in erroneous pressure transient interpretation. These errors can be several orders of magnitude. Elementary AnalysisIn a wellbore where the shut-in tubing pressure is below the dew point, condensate (or water) dropout followed by liquid re-injection is frequently observed. Basically, if there are liquids present in the wellbore in a gas/gas condensate/water system, the gas lifts the liquids when the well is under production. When the well is shut-in, the mechanism by which the liquids are lifted is eliminated (gas velocity ~ 0). Thus, the liquid must either vaporize (fluids become single phase gas) or, since the liquids are much denser than the gas, fall of the well. If the liquids "drop-out", they will then re-inject in to the reservoir, being displaced by much lower density gas. The way that DRC approaches this problem is to perform a flash calculation on the wellbore fluids at the final shut-in condition. This allows us to determine the amount of condensate that can remain in solution with the gas. We then assume that the condensate that is not in solution will fall back and form a liquid column before re-injecting. Unfortunately, while there is a liquid level of unknown height in the well, our bottomhole conversion is not accurate. Fortunately, most moderate to high permeability wells will finish "re-injecting" in a matter of minutes or hours. In case the re-injection period obscures the results of the shut-in, we recommend performing a single choke drawdown following the build-up. Since we can model the behavior of the well while it's flowing, we'll be able to evaluate skin and permeability from the drawdown and P* from the Build-up. Basically, the drawdown can be used to see the portion of the test that was missed during the re-injection.Error AnalysisSince the crux of Pressure Transient Analysis is the observation of CHANGE, any absolute error will have very little effect on the interpretation of permeabiilty or on reservoir volumes. In single-phase wells, there also is no effect on the skin calculation; in multiphase wells, the error is minor. The only parameter that is affected by any error in our BHP conversion is P*; for single-phase wells, the likelihood of having any error in P* is very slight. In any case, we can quantify the amount of error that is possible for each well test objective for a given set of conditions. Our clients may then determine if this error could possibly change the conclusions of the test (high skin? low permeability? small reservoir?). If there are no differences in the conclusions, why bear the expense and take the risk of putting wire and gauges in the wellbore? |
Home About DRC The SPIDR Analysis Tools Literature Contact Us News Careers
13231 Champions Forest Drive, Suite 401,
Houston, TX 77069
Phone 281-444-5398 • Fax 281-444-5397
Questions, comments or suggestions? Contact Us
(c) 2009 DRC - Data Retrieval Corporation
Site by Paradigm Web Design
Site Ensemble CMS Copyright © 2009 Paradigm New Media, LLC. All rights reserved.
