Ann: Preparations underway for start of Zapato-1 exploration well, page-599

Drill pressures and how its done.

Pre-drill pressure prediction

Before drilling, pressure is modeled in the pressure cell (bordered by geological faults) where the well is located. The model is run several hundred times where parameters not known for certain influencing the over pressure are varied slightly. Each simulation results in a pressure prognosis for the well. The mean result as well as the std. deviation is depicted together with the hydrostatic and lithostatic pressure. With greater depths the uncertainty in pressure increases Pressure prediction while drilling.During drilling real time resitivity and sonic data is used for pressure prediction. Using these measurements the pressure is calculated, Equation for Eaton’s method is given below:

Resistivity:

P_obs= s- ( s- p_n) ( R_Sh,obs /R_{sh ,n})^nr

with n_r = 1.2

Equation of pressure from sonic log using Eaton’s method

P_obs= s- ( s- p_n) ( ΔT_Sh,obs /ΔT_{sh ,n})^ns

with n_s = 3.0

s = Overburden stress (Pa) and is given as a function of the average formation bulk density.

p_n = Normal pore pressure (Pa) and is a function of the average formation fluid density.

R_Sh = Shale resistivity (Ùm)

ΔT_Sh,= Shale travel time (s/m)

Note that subscript “obs” refers to the observed (actual pressured) condition and subscript “n” refers to the hydrostatic (normally pressured) condition.

The corrected drilling exponent for normally compacted/pressured formation can be inferred from offset wells at the same depth, or from extrapolation of trend data in the normally compacted/pressured formation.

From the Monte-Carlo runs and the observed pressure calculated from the resitivity, sonic, and drilling exponent can calculate weights w_i given to each simulation run i, based on the log data information. In basin modeling, these weights are calculated based on observed pressures P^obs from a calibration measurement using below equations:
w_i= N / ( ∑a_n (p_n^mod(i) –p_n^obs)²

where N is the number of calibration depths, ‘n’ refers to the well depth, a_n is a weight of importance applied to each calibration depth.
pmod(i ) is the modeled overpressure for depth ‘n’ in run ‘i’, and
pobs is the observed (measured)

overpressure for the calibration depth. An estimator for the most likely value of p is then

p^{= ∑} p^mod(i) w_i/ ∑w_i

In this implementation we calculate pobs as the arithmetic average of the pore pressure calculated according to resistivity, sonic and drilling data. If one of these is missing, it is ignored.