How to QC the PVT data

Well Test Analysis, Pressure and Production Monitoring........ etc.

Moderator: moderators

Post Reply
User avatar
FANARCO
Administrator
Administrator
Posts: 2063
Joined: Mon Jul 10, 2006 9:56 pm
Contact:

How to QC the PVT data

Post by FANARCO »

[align=right]Most of the time for QC purposes in oil PVT you need to perform four basic tests called Density Test, Y Function, Inequality Test and Material Balance Test.[/align]
User avatar
FANARCO
Administrator
Administrator
Posts: 2063
Joined: Mon Jul 10, 2006 9:56 pm
Contact:

Post by FANARCO »

1) Density Test: the density of the oil saturated with gas at bubble point in the differential test should be equal to the density of the oil obtained at the optimum pressure in the separator test. The test is valid of the difference if both densities is not higher than 5%.
The density recombined from the separator test (Pofb) is calculated in the following form:

Pofb = (Mass of oil at tank conditions + Mass of Gas in the separator + Mass of Gas in the tank)/Unit of Oil at Pb and T

Mass of oil = (YoPw)/(Bofb) where Yo is the oil specific gravity, Pw is the water density (lb/STB) and Bofb is Bo at optimum pressure conditions.

Mass of Gas @ Separator = 0.0763277 ((Ygsep Rssep))/(Bofb)
Where Ygsep is the specific gravity of gas and Rssep is the solution GOR, both at separator conditions.

Mass of Gas @ Tank = 0.0763277 ((Ygtan Rstan))/(Bofb)
Where Ygtan is the specific gravity of gas and Rstan is the solution GOR, both at tank conditions.

Example:

Podb = 0.6562 g/cc (oil density fron differential test)
Bofb = 1.474 rb/STB

From optimum separator test pressure:
Separator Conditions Rs = 676 SCF/STB and Yg = 0.786
Tank Conditions Rs = 92 SCF/STB and Yg = 1.363
API = 40.7

Solution:

Pofb = ((141.5/(131.5+40.7))*(350/1.474)) + ((0.0763277/1.474)*((676*0.786)+(92*1.363))
so,
Pofb = 229.12 lb/rb (0.6546 g/cc)

Comparing with the density fron the differential test:

%Dif = (0.6562 - 0.6546)/0.6562
%Dif = 0.24% (Acceptable)
User avatar
FANARCO
Administrator
Administrator
Posts: 2063
Joined: Mon Jul 10, 2006 9:56 pm
Contact:

Post by FANARCO »

2) Y function: using the well know equation Y=(Pb-P)/(P(V/Vb - 1)) you should obtain a straight line or your PVT data must be corrected using some statistical techniques. Most of the time, a deviation from a straight line behavior is due 1. Over or Sub estimation of the bubble point in the lab or 2. Sample contaminated (CO2 for example).
User avatar
FANARCO
Administrator
Administrator
Posts: 2063
Joined: Mon Jul 10, 2006 9:56 pm
Contact:

Post by FANARCO »

3) Inequality Test: is just to compare that dBod/dp < Bg dRsd/dp or a simulation will sent to you a message error. in this case, dBod/dp is the variation of oil volumetric factor (from differential test) with pressure, Bg is the gas volumetric factor and dRsd/dp is the variation of Rs data set from differential test respect to pressure. This test is very straightforward and must be true in all pressure points.
User avatar
FANARCO
Administrator
Administrator
Posts: 2063
Joined: Mon Jul 10, 2006 9:56 pm
Contact:

Post by FANARCO »

4) Material Balance Test: checks if Rs from differential test is the same (or 5% maximum different) to the Rs calculated using material balance. For the Rs calculated, you can suppose one liter of residual oil at standard conditions in order to simplify your calculation.

If your PVT passes this entire test, you can consider it valid. There are some techniques for PVT correction in order to obtain a more accuracy Pb and then extrapolating the data you can obtain a "corrected PVT", but are statistical.

Or simply: Using a pre-processor (for example, PVTp) with the components of your oil, you can simulate the experiment. If the results of the simulated experiment are close to your real PVT, you can consider it valid.
petro
Posts: 35
Joined: Sat Jan 15, 2011 12:02 am

Post by petro »

thanks FANARCO,

this is great subject, as QC of data to be used for reservoir simulation is very important.

Hopefully, more topics like, QC of Kr data, Pc data, and Allocated production data to be discussed.

All the best!
Post Reply