Page 163 - Geothermal Energy Renewable Energy and The Environment

Drilling 149

challenge for equipment, fluids, and operational tools. The high failure rate of equipment at high
temperatures increases the risk of drilling deviated wells, but the benefits still exceed the cost in
many cases.

case sTUdy: kakkonda, Japan

The Kakkonda geothermal field, located approximately 500 km north of Tokyo, Japan, is one of
the world’s shallowest, high temperature geothermal regions. It is located within the active Tohoku
volcanic arc, which is part of the subduction zone complex that bounds the Pacific plate in the west-
ern Pacific region. In the Kakkonda region, very young granitic rocks, dated between 110,000 and
240,000 years old (Doi et al. 1995; Matsushima and Okubo 2003) intruded to depths of less than 3
km, and supply a prodigious quantity of heat to the area.
More than 70 geothermal wells have been drilled in the area, most reaching depths of between
1000 and 3000 meters. The temperature distribution as a function of depth is variable (Figure 8.10).
Power has been generated in the area since 1978—currently 78 MW are generated in the region
(Matsushima and Okubo 2003).
To better delineate the nature of the geothermal resource and to evaluate various drilling tech-
nologies and strategies, a deep well, designated WD-1 (later WD-1a) was drilled in the region in
1994 and 1995. The target depth was 4000 meters. A head drive drilling method was chosen to
evaluate, among other things, the ability of such a system to provide adequate cooling during drill-
ing. A mud cooling system was also employed.
The history of events during that drilling program provides insight into the challenges to be
expected drilling in such a setting. We will consider these in the following discussion, which is
based on events summarized from Uchida et al. (1996) and Saito and Sakuma (2000).
Drilling was initiated on January 5, 1994. The approach was to redrill an earlier hole that had
been drilled in 1992 and 1993. The plan was to use a five casing design, similar to that depicted in
Figure 8.8. The depths, diameters of drill bits, and diameters of the emplaced casing are shown in
Figure 8.10.
Previous experience from drilling geothermal wells in the region indicated that the volcanic rock
in the region was highly fractured. For this reason it was anticipated that lost circulation would be
an important challenge when drilling the deep hole. Of particular concern was the region where
the shallow geothermal reservoir existed. In fact, lost circulation was a major challenge throughout
the drilling of the first 2000 meters (Figure 8.10). Cement was successfully used to close off many
of those zones, but eventually the well trajectory had to be modified at a depth of about 1,696 m
because of this problem.
In the region where lost circulation was most common due to the intense fracturing of the
rock, the average rate of penetration was approximately 6.7 m/day. Beyond that zone, the pene-
tration rate was almost 18 m/day. Although common experience under “normal” conditions sug-
gests that 50% of the cost and time will be committed to completion of the last 10–20% of the
hole, this example demonstrates unequivocally that such a rule of thumb is highly conditional.
The expectation at the time of drilling was that the high temperature geothermal reservoir would
be encountered at about 3000 m depth. Although temperatures in excess of 200°C were encountered
initially at that depth, there was no indication of the expected high temperature reservoir so drilling
was continued.
At a depth of 3350 m, elevated CO contents in the drilling mud were detected. Elevated CO can
2
2
be a health hazard and needs to be controlled. Lime was added to the mud to reduce the CO , with
2
the intent of capturing the CO as calcium carbonate. This effort appeared to be successful.
2
At depths beyond 3451 m, it was discovered that the drilling mud deteriorated because of high
temperatures when mud pumping was stopped in order to remove the drill string and change the bit.
3
To overcome this problem, the mud density was lowered to less than 1.1 gm/cm . It was anticipated
that this lower density and higher water content would provide additional time to allow bits to be