Geothermal drilling cost and drilling effectiveness.

Abstract

The cost of geothermal wells and field development is about 40% of the total investment cost for new high-temperature geothermal plants. This “up front” cost makes geothermal plants more expensive to build than conventional plants and because of this and the perceived risk, a lot of attention has been focused on ways to reduce this cost. This paper describes the drilling cost structure and what factors affect the cost. About half of the well cost is related to the time charges of the drilling rig (day rates) and associated equipment and thus ways of reducing the time it takes to drill the well is one way of reducing the overall cost. There is surprisingly little published data available on the breakdown of geothermal drilling costs, because of the competitive nature of the drilling market and confidentiality clauses.

Full text

Presented at “Short Course on Geothermal Development and Geothermal Wells”,
organized by UNU-GTP and LaGeo, in Santa Tecla, El Salvador, March 11-17, 2012.
1
LaGeo S.A. de C.V.
GEOTHERMAL TRAINING PROGRAMME
GEOTHERMAL DRILLING COST
AND DRILLING EFFECTIVENESS
Sverrir Thorhallsson1 and Bjorn Mar Sveinbjornsson
1Iceland GeoSurvey (ÍSOR),
Grensásvegur 9, 108 Reykjavík,
ICELAND
sverrir.thorhallsson@isor.is, bms.isor@gmail.com
ABSTRACT
The cost of geothermal wells and field development is about 40% of the total
investment cost for new high-temperature geothermal plants. This “up front” cost
makes geothermal plants more expensive to build than conventional plants and
because of this and the perceived risk, a lot of attention has been focused on ways
to reduce this cost. This paper describes the drilling cost structure and what factors
affect the cost. About half of the well cost is related to the time charges of the
drilling rig (day rates) and associated equipment and thus ways of reducing the
time it takes to drill the well is one way of reducing the overall cost. There is
surprisingly little published data available on the breakdown of geothermal drilling
costs, because of the competitive nature of the drilling market and confidentiality
clauses.
1. INTRODUCTION
Drilling performance of 73 high temperature and reinjection wells drilled in the period 2001–2009 in
the Hengill field in Iceland was analysed and the statistical level of risk assessed (Sveinbjornsson,
2010). The following paper reports the main topics of that reference. The number of working days to
complete each depth section of the well (4 sections) was analysed and the time broken down to show
how much was spent on drilling, tripping, casing, cementing, logging etc. The results were then
grouped according to which design was used and technology applied. Cost calculations were made,
based on market prices, as the real cost was not made available. The time breakdown had similarly to
be worked out from the geological reports as the key performance indicator (KPI) data was
confidential.
2. TIME ANALYSIS
The most common type of high temperature well in the Hengill field is of a “large diameter” casing
program. The drilling is divided into four sections. The initial pre-drilling (Section 0) is by a small
rig (50 t) with a 26" bit down to 90 m for a 22½" surface casing, followed by Section 1 drilled by a
larger rig (200 t) with a 21" bit to 300 m for the 18⅝" anchor casing, Section 2, inclined or directional
drilling with a 17½" bit to 800 m for 13⅜" production casing, and Section 3 with a 12¼" bit to a depth
of 1,800 to 3,300 m for 9⅝" slotted production liner. To compare the drilling time for different wells

Thorhallsson and Sveinbjornsson 2 Drilling cost and effectiveness
the respective numbers of workdays were normalised for a reference well of that design and the
average depth of the group which was 2,175 m. Table 1 shows the results.
TABLE 1: Normalised workdays for large diameter reference wells
Drilling Project
Workdays total
Beta-PERT
Section
Drilled
Number
Lowest
Most
likely
Highest
Average
Standard deviation
(m)
(n)
(a)
(m)
(b)
(te)
(s)
%
0
90
23
3
5
11
5.7
1.3
22.8
1
210
35
5
8
14
8.5
1.5
17.7
2
500
48
6
10
18
10.7
2.0
18.7
3
1,375
50
10
16
38
18.7
4.7
25.1
Total
2,175
43.5
5.5
12.6
The number of wells varies as fewer reports were available on the sections of pre-drilling and drilling
for the anchor casing than the sections of drilling for the production casing and the productive open
hole. Out of the 73 wells, drilling of a section in 14 wells ran into unusual difficulties which led to an
excessive number of workdays and increased material cost. The difficulties were mostly due to
anomalous geological conditions. As excessive cases of this nature would skew the distribution of
normal drilling progress it was decided to analyse the frequency of them separately but exclude from
the reference class six of these cases which deviated more than 3 standard deviations from the average.
Figure 1 below shows the distribution of the resulting reference class for the total of workdays in
normal drilling of large diameter wells. The distribution is of the asymmetric Beta-PERT type where
the most likely value is lower than the average. With 95% confidence the workdays lie between 33.4
and 55.5 days. The average for the empirical data of the total is 43.5 days.
FIGURE 1: Distribution of total workdays for large diameter reference wells to 2,175 m
The workdays were also analysed for each section of drilling and the time used for different activities
such as actual drilling, running and cementing casing, delays due to drilling problems, logging, install-
ation of wellhead and other reasons for delays. The results of that analysis are shown in Table 2.
Workdays

Drilling cost and effectiveness 3 Thorhallsson and Sveinbjornsson
TABLE 2: Workdays of different activities for large diameter reference wells
Holes
Workdays total
Workdays of different activities
Section
Drilled
Number
Average
St.
dev.
Drilling
Casing
Problems
Logging
Completion
Other
(m)
(n)
(te)
(s)
(d)
(d)
(d)
(d)
(d)
(d)
0
90
23
5.7
1.3
2.4
1.9
0.4
0
0.9
0.1
1
210
35
8.5
1.5
3.0
2.2
1.0
0.8
1.1
0.4
2
500
48
10.7
2.0
5.0
2.4
0.5
1.1
1.3
0.4
3
1,375
50
18.7
4.7
9.4
1.1
1.9
4.0
1.6
0.7
Total
2,175
43.5
5.5
Besides the analysis for the reference well of the “large diameter” type it is of interest to compare the
number of workdays to that of the “regular diameter” casing program of with casing diameters of
18⅝" surface, 13⅜" anchor, 9⅝" production casing and a 7" slotted liner. The results are shown in
Table 3. The number of wells varies according to the availability of reports. The average and the
standard deviation are calculated assuming a Beta-PERT distribution for the workdays. The total
workdays for the regular diameter wells are 46.9 days but 44.1 days for the large ones. The difference
is insignificant, but if any it takes slightly less time to drill the large diameter wells.
TABLE 3: Workdays for regular and large diameter wells to 2,175 m
Section 0
Section 1
Number of
holes
Work-
days
St. Dev.
(s)
Number of
holes
Work-
days
St. Dev.
(s)
Regular diam.
17
6.3
2.6
21
8.7
3.8
Large diam.
23
6.1
2.2
35
8.4
2.3
Section 2
Section 3
Total
Number of
holes
Work-
days
St. Dev.
(s)
Number of
holes
Work-
days
St. Dev.
(s)
Work-
days
St. Dev.
(s)
Regular diam.
20
10.6
4.2
22
21.3
7.6
46.9
9.8
Large diam.
48
10.5
2.8
50
19.1
6.6
44.1
7.8
3. COST ANALYSIS
The cost structure is such that there is a day rate for the drilling rig and crew and also for the many
services engaged such as for cementing, directional drilling, drilling mud, logging etc. These daily
costs vary according to the technology requirements of the equipment, geographic area, and prevailing
market conditions. The unit material costs on the other hand reflect the commodity prices for steel,
cement, fuel oil etc. and their overall cost is therefore more predictable as the usage quantity can be
calculated. On top of this the remoteness of the site and proximity to supplies and services affect these
costs.
The cost of drilling the reference well of the large diameter program was calculated on the basis of the
number of workdays required for each section of the drilling. A breakdown of cost for different
sections is shown in Table 4.

Thorhallsson and Sveinbjornsson 4 Drilling cost and effectiveness
TABLE 4: Average cost of a large diameter reference well to 2,175 m
Component
Cost ($)
(%)
Site, cellar, water supply
400,000
8.6
Moving in the smaller drill rig
106,000
2.3
Moving in the larger drill rig
255,000
5.5
Site and Moving total
761,000
16.3
Section 0, small rig, 26” to 90 m for 22½” surface casing
332,000
7.1
Section 1, large rig, 21” to 300 m for 18⅝” anchor casing
716,000
15.3
Section 2, large rig, 17½” to 800 m for 13⅜” production casing
1,303,000
27.9
Section 3, 12¼” to 2.175 m for 9⅝” slotted production liner
1,556,000
33.4
Total
4,668,000
100
It is assumed that a small drill rig is used for Section 0 to 90 m, but the Sections 1, 2 and 3 are drilled
by a larger rig. Note that this is the average cost of a large diameter reference well where unusual
problems in 6 wells that led to deviations in excess of 3 standard deviations from the average have
been excluded. The risk of such problems was dealt with separately (Figure 6). Table 5 shows the
time cost and material cost for each section as well as the percent of the total cost.
TABLE 5: Breakdown of cost of a large diameter reference well to 2175 m
Item of
Cost
Section 0
Pre-
Drilling ($)
(%)
Section 1
Anchor ($)
(%)
Section 2
Pro-
duction ($)
(%)
Section 3
Open hole
($)
(%)
Total
($)
(%)
Time cost
total
172,665
52.0
430,193
60.1
569,560
43.7
1,093,825
70.3
2,266,243
35.2
Material
cost total
159,424
48.0
285,777
39.9
733,375
56.3
462,918
29,7
1,641,494
48.4
Site etc.
400,000
8.6
Moving
small rig
105,625
2.3
Moving
larger rig
255,000
5.5
Total
332,089
100
715,970
100
1,302,935
100
1,556,743
100
4,668,362
100
4. REFERENCE CLASS FOR THE TOTAL COST
To obtain an estimate of the variance in total cost Monte Carlo simulations were carried out using
probability distributions for the uncertainties in the number of workdays, the unit costs of material,
and day rates for the drilling rigs. Figure 2 shows the distribution for the total cost of the reference
well of the large diameter program. Note that here the cost of the drill site, cellar and water supply, as
well as the cost of moving rigs in, are included. The average obtained for the simulation is
$4,665,000, compared to the total cost of $4,668,000 obtained in Table 5. The standard deviation was
found to be $359,200. The cost lies with 95% confidence within the limits $4,101,000 and
$5,365,000. Sensitivity analysis shows that the number of workdays causes most of the uncertainty,
76.5% in Section 3, 13.4% in Section 2, 5.9% in Section 1 and 1.9% in Section 0. Graphs for
accumulated probability indicate that in 30% cases the cost exceeds $4,825,000 and in 30% cases the
cost will be lower than $4,457,000.

Drilling cost and effectiveness 5 Thorhallsson and Sveinbjornsson
FIGURE 2: Total cost of the large diameter reference well to 2,175 m
5. WORKDAYS AND COST WITH DEPTH
Figure 3 shows the depth of a large diameter reference well as a function of the number of workdays.
Note that workdays for moving drill rigs in are included here. The work components drilling and
problems are counted as active drilling time but moving in, setting up the rig, casing, cementing,
logging, well completion and other are counted as waiting time.
FIGURE 3: Depth of the large diameter reference well versus number of workdays
-2.500
-2.000
-1.500
-1.000
-500
0
0 5 10 15 20 25 30 35 40 45 50 55
Depth (m)
Workdays (d)

Thorhallsson and Sveinbjornsson 6 Drilling cost and effectiveness
Figure 4 shows how the cost of a large diameter reference well increases with well depth. The cost of
drill site, cellar, water supply, and moving rigs in, is included here. This graph is useful in estimating
the cost of each section and what would be lost if section 3 must be redrilled. Also how much it would
cost to deepen the well beyond the depth of a reference well.
FIGURE 4: Cost of the large diameter reference well versus depth
Figure 5 shows how the cost of the well increases with the number of workdays.
FIGURE 5: Cost of large diameter reference well versus number of workdays
6. UNUSUAL DRILLING PROBLEMS
In the Hengill field 14 of the 73 wells encountered unusual problems which led to additional cost. The
main causes were difficult geological formations where the bit got stuck in the hole. In 6 cases or
8.2% the additional cost exceeded 3 standard deviations of the reference class (3 x $359,200).
0
500.000
1.000.000
1.500.000
2.000.000
2.500.000
3.000.000
3.500.000
4.000.000
4.500.000
5.000.000
0
100
200
300
400
500
600
700
800
900
1.000
1.100
1.200
1.300
1.400
1.500
1.600
1.700
1.800
1.900
2.000
2.100
2.200
2.300
Cost ($)
Depth(m)
0
500.000
1.000.000
1.500.000
2.000.000
2.500.000
3.000.000
3.500.000
4.000.000
4.500.000
5.000.000
0 5 10 15 20 25 30 35 40 45 50 55
Cost ($)
Workdays (d)

Drilling cost and effectiveness 7 Thorhallsson and Sveinbjornsson
Figure 6 shows the number of wells where the additional cost due to unusual problems exceeded
$250,000. This distribution can be of aid in estimating additional risk due to such unusual problems
on top of the risk included in the statistical distribution of the reference well.
FIGURE 6: Number of wells with additional cost due to unusual problems exceeding $250,000
7. POWER OUT OF WELLS
The overall economics of a geothermal power project is strongly influenced by the power output per
well, or how much can be reinjected, which is also considered in evaluating the drilling effectiveness.
Table 6 shows the power output per drilled geothermal well and per productive well in the Hellisheiði
region of the Hengill field. It is of interest to note that the difference between the regular and large
diameter wells appears insignificant.
TABLE 6: Power output of wells
Diameter
Drilled wells
Productive
wells
Power per
drilled well (MWe)
Power per productive
well (MWe)
Large diameter
38
33
5.8
6.7
Regular diameter
15
13
5.7
6.6
Total
53
46
5.8
6.7
The data bank could be used for other comparisons such as vertical vs. directions wells, drilling with
water only or managed pressure drilling by aerating the water. Only 4 of the wells in the Hengill field
were however drilled vertical. A comparison with vertical wells is therefore not reliable. For success
metrics, comparisons were made between the Injectivity Index (II) at the end of drilling and the
confirmed flow-output (MWe or kg/s of steam and water) of the well. The results indicate that to
obtain reliable predictions of yield on the basis of the Injectivity Index one must also consider
reservoir conditions and enthalpy of the expected discharge. Such predictions would be valuable for
decisions, whether to deepen a well or redrill the last section as a sidetrack or “fork”.
0
2
4
6
8
10
12
14
16
0 1.000 2.000 3.000 4.000 5.000
Frequency
Cost ($1.000)

Thorhallsson and Sveinbjornsson 8 Drilling cost and effectiveness
8. COMPARISON BETWEEN ICELAND AND KENYA
Thomas Miyora Ongau, a UNU fellow in 2010, compared the time required to drill 12 directional
wells from Kenya to 14 similar wells of regular diameter from Iceland. These selected wells have the
same casing sizes but the Kenyan wells are deeper (Table 7). The wells have 9⅝” production casing
and are directionally drilled to total depth with an 8½” bit. The Iceland wells are a subset of the 22
regular diameter wells analysed above and the time data for the Kenya wells is from drilling records
and recorded KPI’s.
TABLE 7: Depths of wells in Kenya and Iceland, regular diameter (Miyora, 2010)
Kenyan wells
Icelandic wells
Steps
Depths (m)
Steps
Depths (m)
0
0-60
Surface casing
0-90
1
60-300
Anchor casing
90-300
2
300-1000
Production casing
300-800
3
1000-2800
Production liner
800-2300
Figures 7 and 8 show the results of a breakdown of drilling time for the Kenyan and Icelandic wells.
Table 8 shows percentages of drilling time for both groups of wells. In Kenya a greater percentage is
spent in drilling and changing of bits whereas relatively more time is spent on logging, cleaning and
casing in Iceland.
The overall advance from start to finish of the drilling is about 58 m/day at Hengill vs. 48 m/day for
Kenya. The workdays required to drill the average depth of the Kenyan wells of 2.767 m were 57.3
days in Kenya but 49.2 days in Iceland.
In both countries about 80% of the wells were drilled according to plans but 20% were “problem
wells” mainly due to the rigs getting stuck and other geological risks.
TABLE 8: Breakdown of drilling time in percentages for similar wells in
Kenya and Iceland (Miyora, 2010)
Kenya
Drilling
Casing
Cem.
Plug
Stuck
Ream
Fish
Kenya
57.94
4.42
7.40
0.47
1.26
3.22
0.42
Iceland
45.31
8.33
5.29
4.45
4.99
2.16
0
Kenya
Water
bit/BHA
Repair
Cleaning
Logging
Other
Kenya
0.37
9.55
2.02
1.66
4.93
6.35
Iceland
0.12
0.95
1.16
9.43
17.52
0.28

Drilling cost and effectiveness 9 Thorhallsson and Sveinbjornsson
FIGURE 7: Time analysis for regular diameter wells in Kenya (Miyora, 2010)
0
10
20
30
40
50
60
70
80
Days
Other
Measurement
Hole cleaning
Repair
Change bit/BHA
Wait on water
Fishing
Reaming
Stuck
Cement plug jobs
Cement
Casing
Drilling

Thorhallsson and Sveinbjornsson 10 Drilling cost and effectiveness
FIGURE 8: Time analysis for regular diameter wells in Iceland (Miyora, 2010)
REFERENCES
Miyora, T.O., 2010: Controlled directional drilling in Kenya and Iceland. Report No. 20 in:
Geothermal Training in Iceland 2010. UNU-GTP, Iceland, 365-390.
Sveinbjornsson, B.M., 2010: Cost and risk in drilling high temperature wells in the Hengill field.
MSc thesis, University of Iceland, 62 pp. + appendices.
0
10
20
30
40
50
60
70
80
Other
Measurement
Hole cleaning
Repair
Change bit/BHA
Wait on water
Fishing
Reaming
Stuck
Cement plug jobs
Cement
Casing
Drilling