Page 72 - Practical Well Planning and Drilling Manual

Section 1 revised 11/00/bc 1/17/01 2:56 PM Page 48

[ ] Well Design
1.4.4

If rafts of other material can be seen within a massive salt, the well-
bore may be moved to avoid them. Unfortunately, the sonic qualities of
massive salts make it hard to interpret features within it to give an
accurate assessment of the location. Where rafts are not frequent the
margin of error could be accounted for. In some cases, it may not be
possible to completely avoid any chance of drilling into a raft, with the
attendant danger of a high-pressure kick. In this case the best that may
be done is to ensure that casing is set as low as possible before reach-
ing the critical depth so that at least a good shoe strength will be avail-
able if the well kicks.
Interpreting seismic is outside the scope of this book. The drilling
engineer planning a well should spend time with a competent seismic
interpreter and ensure that the proper questions are asked to obtain
the fullest picture of the subsurface features.
Examining offset well records for overpressure prediction. If a planned
well is in an area where offset wells have been drilled, these should
provide good data for flagging areas of abnormal pressure. Offset data
points should be plotted and marked with the well name on a depth-
pressure graph; these will give pore pressure (PP) and fracture gradi-
ent (FG) trends for the offset wells, which should give an indication
(when combined with overall field knowledge) of the likely trends in
the planned well. In some cases a full pore pressure-frac gradient
analysis may have been done and, if this is felt to be reliable, can be
plotted directly on the graph.
As with all other data, try to evaluate raw data rather than relying
on other people’s interpretations. Mistakes may have been made in the
earlier interpretation that could then mislead you. This does happen!
Direct measurements such as RFT or MDT pressures provide very
accurate and reliable pressure data points and formation fluid gradi-
ents. These can then act as a qualitative check on other data. If a kick
is taken then the formation pressures will be well known, and this also
should be reliable data.
If total losses were taken and the well was filled with water to find
a static level, then the pressure in the loss zone can be calculated if the
depth is known. It is commonly assumed that if losses are suddenly
taken while drilling that the loss zone is on bottom, but this is not nec-
essarily the case. Before you rely on such data, try to establish whether
another potential loss zone was exposed higher up in the open hole.
Just as pore pressures and fracture gradients are intimately related,
frac data (from leakoff tests or frac jobs) provide data points that can

48