PROCEEDINGS, Thirty-Ninth Workshop on Geothermal Reservoir Engineering

Stanford University, Stanford, California, February 24-26, 2014

SGP-TR-202

Aerated Underbalance Drilling Screening Assessment

at “X” Geothermal Field in Indonesia

Ariya Dwinanto and Sudjati Rachmat
Geothermal Lab. Gd. Teknik Perminyakan FTTM - Institut Teknologi Bandung
Jalan Ganesa 10, Bandung 40132 Indonesia
[email protected]

Keywords: underbalance drilling, screening; aerated drilling

ABSTRACT
Geothermal fault network often has low pressure or subnormal pressure with high temperature condition (>225oC). Low pressure
results to a loss circulation problem. Underbalance drilling method has an advantage dealing with low pressure reservoir.

This paper introduces an aerated underbalance drilling screening method. This study will help the quantitative and qualitative
decision whether the underbalance drilling is feasible or not. The first phase qualitative decision is based on wellbore stability, loss
circulation, reservoir damage, stuck pipe incident, hard drilling and cost benefit. Then it goes to the drilling fluid decision and a
quantitative decision for constructing a feasible bottom hole pressure window area with hole cleaning assessment. Underbalance
drilling assessment is studied on field “X” at one of Indonesia’s geothermal field. The conclusions are that it has opportunities.
Underbalance drilling can be implemented with vertical aerated drilling wells on specific gas and liquid flow rates.

1. INTRODUCTION
Indonesia has good resources of convective hydrothermal geothermal system, which heat transfers mostly on fluid media. Those
fluid media exist on a vapour dominated or liquid dominated phase. Fluid moves on fault network that is the main target of drilling
in geothermal. Fault network gives the conduit facility for recharge fluid to flow in to the heat source as recharge zone and fault
network also gives the conduit facility for formation fluid to emerge to the surface as up flow / outflow zone.

Fault network is a challenging problem for drilling operations. Formation fluid contains high temperature production fluid which
can reach >225oC on high enthalpy system. The other consequecies is that almost all fault network has low pressure or subnormal
pressure (Hole, 2006). This low pressure results to a loss circulation problem. High temperature fluid and low fomation pressure are
main problems for geothermal drilling.

Conventional overbalance drilling has weaknesses if considering low pressure reservoir. Drilling fluid pressure is intentionally
higher than pore pressure. As a result, loss circulation becomes a common problem. Underbalance drilling method has an advantage
dealing with low pressure reservoir. Loss circulation problem can be mitigated by the drilling fluid pressure which is intentionally
lower than the reservoir pore pressure.

Underbalance drilling with it’s advantages and weaknesses has to be studied. This study will help the quantitative and qualitative
decision whether the underbalance drilling is feasible or not. It will help to decide which drilling fluid is compatible. This study will
construct a feasible bottom hole pressure window area with some hole cleaning assessment.

2. METHODOLOGY
The two first process (from figure 1 ) will give the fluid fraction and production casing diameter (big hole or small hole) input data
which are needed for process three. Process reservoir characterization determines the well type, based on production fluid and
reservoir domination type. Process casing diameter feasibility with wellbore stimulation determines the casing diameter.

Underbalance screening include decisions whether underbalance has an opportunity or not and drilling fluid selection.
Underbalance screening is studied on literature and is compared with the actual field offset drilling events. Al-ajmi (2003) gives a
great method for the process. Unfortunately some of the screening process need expert judgements. These subjective judgements
will be mitigated by some graphic based on Indonesia’s geothermal data. Thus people with less experiences may extract data for
parameters at a same results.

After the debit approximation is determined, we can go to the next step. Bottom hole pressure will be counted. Bottom hole
pressure will effect the drilling fluid properties. And the drilling fluid properties will effect the bottom hole pressure, thus iterations
are needed. Lyon et. al. (2009) describe underbalance techniques to determine bottom hole pressure.

One of important parts of drilling operation is the circulation system. The fluid has to guaranteed to lift cuttings up to the surface.
Lifting cutting often known as hole cleaning process. In an effective hole cleaning process, there is a minimum velocity that should
be implemented. The minimum velocity will be counted using Lyon’s et. al. (2009) method. Bottom hole pressure and hole
cleaning process will give two boundary criteria needed to conduct bottom hole pressure window area. Furthermore to conduct the
bottom hole pressure window, we have to study the boundary. The boundary limitation is considered from : bottom hole pressure,
liquid phase drilling fluid flow rate, vapor phase drilling fluid flow rate, motor pump capacity, reservoir pressure drop and hole
cleaning assessment. These consideration will describe the bottom hole pressure window area.
1
Dwinanto and Rachmat

Start Start

Input production
test data
Reservoir characterization
Wellbore simulation

Casing diameter feasibility

Wellbore simulation for
with wellbore simulation standard & big hole

Generate flow rate from

Underbalance screening results
process

Generate casing properties

Bottom hole pressure

calculation
Velocity calculation

Hole cleaning assessment V > 2.44 m/s

No Yes

Bottom hole pressure Approved Bad design

window area determination

End End

Figure 1 (Left side) Methodology flow chart; Figure 2 (Right side) Casing feasibility flow chart
2.1 Reservoir characterization
Each reservoir has it’s own characteristic. Geothermal reservoir can be distinguish by the production fluids. There are : water
dominated reservoir or vapor dominated reservoir. Using boiling per depth graphic, a well is considered single phase or double
phase. Then each of the wells data are gathered to give a conclusion about what is the reservoir type.

2.2 Casing diameter feasibility

To compute wellbore simulations, a set of production test has to be known. Then a wellbore simulation may be conducted to
achieve a simmillar behaviour such as known on the datas. Wellbore simulation gives a prediction of the well ability (productivity
index / PI) to produce at a certain debit with a specific pressure. The production simulation needs the fluid enthalpy that describe
the quality of the production fluid. These data is simulated to conduct the well pressure behaviour as similar as possible. Afterward
a standard hole and a big hole simulation based on the productivity index prediction can be conducted. These processes will
generate flow rate datas which is usefull for the next step.

Analysing production casing geometry will give an area that a production fluid may pass through. Casing catalogue will give
various of casing inner areas. Casing that had been choosen then are computed to give a velocity for a flow rate generated from
wellbore stimulations. Equation 1 generates velocity for certain value of flow rate.

(1)

An upper velocity limit should be considered as max limit. As above this limit, noise and erosion will take place. Lyle (1947) had
predicted the upper limit for a one phase liquid flow in a steel pipe tube which is 2.44 m/s. Equation 1 with process number one
until five give a velocity prediction. Thus will give a clue about what diameter should be implemented.

3. UNDERBALANCE SCREENING PROCESS

Underbalance candidate screening process is a broad assessment. It will need a great span of data and a great comparison for it’s
litrature riview. Al-ajmi (2003) had been researching and studying many literature. He sugested two screening process subsequently
: underbalance screening and drilling fluid screening process. He came with a sequence of methodology that can be used on mainly
oil and gas reservoir. With some adjustment for geothermal condition, this method can be a powerfull method to decide to go with
underbalance method or not on a geothermal reservoir. Then after a yes decision is made, it goes thru a drilling fluid assessment
which will guide user to use proper drilling fluid for such reservoir. The assessment sequence will be explained by subject briefly.
All of the entitiy process on the flowchart are not mandatory. it depends on the flowchart guide that is clear on figure 5.

3.1 Stable rock anticipated

Wellbore stability is important. The ability of the formation to support and withstand the hole will be good if the hole does not cave
in, collapse or getting narrower. Those problems are going to cause issue such as pipe sticking or even drilling failure which can
incentifies drilling non performing time.

Wellbore instability problems may exist from these factors:

 The formation contains significant amount of water sensitive clays.
 The formation is weak.

2
 Dwinanto and Rachmat

Clays that is prone to swelling are clay which are the most water sensitive clays. To determine clay swelling potential, it is assumed
that the clay types, volume fractions, and distribution are known. Vitthal et al (1989) introduces clay swelling index. The index
given on equation 2, predicts how bad a formation clay can swell.

(2)

Table 1 Damage indices for pure clays

Clay Swelling Fine-migration
component Index (IS) Index (IF)
Montmorillonite 10 10
Chlorite 1 6
Kaolinite 1 6
Illite 2 8
Vermiculite 5 2

Table 2 Distribution correction factors

Swelling Fine-migration
Clay distribution
Factor (CS) factor (CF)
Pore lining 1 0,7
Pore filling 1 1
Pore bridging 0,5 1
Discrete
0 0,9
participles
Thin lenses 0,5 0

Prone swelling clays are: montmorillonite, chlorite, kaolinite, illite and vermiculite. Clays react differently when it counter water.
The most susceptible clay is montmorillonite, thus it’s swelling index indicates value 10 which is the highest swelling index value.
Chlorite and kaolinite at the other side give the lowest swelling index value which is 1. These swelling index are calculated with the
clay distribution factor, thus give a swelling potential govern by equation 2.

There is the upper limit of density so that underbalance drilling can exist and the lower density limit so that the wellbore can with
stand the strain to achieve wellbore stability. The lowest limit to be considered for mud weight should be borehole collapse pressure
gradient. This borehole collapse pressure gradient is computed by uncontrollable factors: local borehole stress, pore pressure and
rock strength conditions. Also controllable factor will help maintaining stability: wellbore fluid pressure. Another consideration is
on hole angle. As the hole angle increases, the borehole collapse pressure gradient will decrease. Controllable and uncontrollable
factor will effect the lower density limit. (Fuh, 1988)

3.2 Lost circulation

Lost circulation is the event when a significant amount of drilling mud is lost to the formation. It occurs when natural, or induced
formation openings are large enough to allow mud to pass through. Al-Ajmi (2003) determines a qualitative method to assess lost
circulation potential. The lost circulation potential is based on two features on a reservoir which is fractures and vugs. The fracture
( enhance significant on rock permeability. Vugs ( is a small cavity in a rock or vein, often with mineral lining of different
composition from that of the surrounding. It is much narrower if considering spaces in fractures. An index ranging from 1 to 10 for
each fractures and vugs is introduced. Other two features which is based on a quantitative estimate to assess lost circulation are The
permeability (k) and the porosity (ф). Equation 3 describe how a lost circulation index can be computed.

(3)

Guideliness which determine lost circulation problems :

 5 % have severe lost circulation problems
 0.1 % have no lost circulation problems
 0.1% 5% has some losses.

Al-Ajmi (2003) explains that only an expert can use equation 3. This may limit it’s value. Thus to make it more usefull, a fracture
graphic based on Indonesia’s geothermal well lost circulation events is made. Figure 10 illustrates that by knowing a average
number of lost circulation events on a well or a few well, the fracture index can be generated. Vugs index can be generated from
figure 11 and equation 4. The effective porosity indicates the closest meaning to a quantitative vug index. Effective porosity value
from 2335 stones had been quantified by Mosch and Siegesmund (2007).

(4)

3.3 Reservoir damage

Formation damage is a term when a reservoir has reduction of permeability. The effect will impact well productivity. Well
maximum ability to produce fluid will be hampered because of positif skin factors. Overbalance drilling as it is mentioned, makes a
big posibility for particles of drilling fluid to go thru filter cake. After the particle go thru the filter cake barier, it may interact with
materials imbedded at the formation. To quantify formation damages, only fine migration and clay swelling will be used.

3
Dwinanto and Rachmat

Vitthal et al (1989) clay swelling index and fine migration. The index given on equation 2, predicts how bad a formation clay can
swell. The index fine-migration index on equation 5 guides to a value which describe the fine migration problems. These indices
estimate the probability for particular formation damage. These indices are assigned to each clay on the basis of its distribution in
the rock, its origin (authigenic or detrail), and its composition. Fine migration index uses table 1 and table 2 for input datas.

(5)

3.4 Stuck pipe

Overbalance method have a difference with underbalance method if considering pipe sticking. The important thing that differs from
those drilling method is only differential pipe sticking. Thus differential pipe sticking has to be quantified. Sharif, Q (1997)
introduces a probability function that pipe sticking will happen or not. Sharif’s differential is described 1 as a definitely must
happen and 0 as cannot happen. This study is based on 143 wells drilled offshore in Uni Emirate Arab. The probability function for
a particular well is determined by equation 6 and equation 7.

(6)

(7)

3.5 Hard drilling

Hard drilling conditions are experienced when drilling through dense formations with low permeability and low porosity. When
hard drilling, it has been reported an increase in penetration rates of up to ten-fold in these formations when drilling underbalanced
(Al-ajmi, 2003). Al-ajmi unfortunately did not give a objective way to quantify and classify those conditions.

The rather effective and objective way to classify a dense formation is to see these figure 8 below. Graphic from figure 8 classifies
rocks based on their properties into low and high properties. For permeability properties, there are not enough data to conclude low
and high properties. For practical reason a graphic (figure 10) had been made based on lost circulation event experienced by wells.
This graphic is the same graphic to quantify lost circulation index. Porosity classification can be seen and generated from figure 9.
Porosity and rock density value are generated from figure 11. These three graphics are assumed to classify rocks thru through dense
formations with low permeability and low porosity (or hard drilling).

3.6 Cost benefit

Underbalance drilling (UBD) as present is a high end technology. This will result to bigger cost regarding the additional
specification of equipment and additional cost for UBD personnel. At the other hand UBD offers a great deal of benefit regarding to
reducing drilling operation cost, bigger reservoir productivity, and earlier production. Spesificaly on geothermal, earlier production
can not be maintained since it is a power generation activity. Thus any UBD planning has to compare these benefits and losses. The
aim is definitely having a bigger benefit trade off if comparing with overbalance drilling project.

Al-ajmi (2003) said additional cost on UBD related to equipments and personnel. Underbalance drilling uses air to decrease the
density of the mud. This characteristic makes an obligation for UBD to install additional tools compared with overbalance method.
Additional tools used in safe UBD include: rotating control head, diverter/rotating BOP, multiphase separators,
compressors/boosters, nitrogen membrane unit, and specialized personnel.

A significant reduction on drilling time may add the benefit of UBD. UBD has been known to increase rate of penetration (ROP)
from 2 up to 3 fold times faster. This faster drilling benefit is accelerated when minimizing non-productive time (NPT) can be
implemented (example: reducing differential pipe sticking problem). Other benefit will be bits usage for 2 times longer comparing
with conventional drilling method. And the last benefit is UBD prevent the loss of expensive drilling fluid. Faster ROP, minimizing
NPT, longer bit life and expensive drilling fluid lost prevention may add to the benefit of UBD. UBD can minimize formation
damage, thus resulting a better productivity index. On a long span of production period, a biggier productivity index result to a
much better income. Al-ajmi (2003) resume that increase ROP, minimizing NPT, longer bit life, minimizing costly-drilling fluid
lost and minimizing formation damage are benefits that will gain less production cost and much more income.

3.7 Production fluid characterisation

Production fluid characterisation data needs water fraction / vapor fraction and H2S gas data. These data will comply to figure 6 and
figure 7 to determine which drilling fluid is compatible for certain situation. Water fraction can be known from production test
results. These data include enthalpy which is usefull for detemining what quality of vapor do the production fluid has. The bigger
the enthalpy, the greater chance it is to have a better vapor fraction. H 2S gas data is deliberated on geochemistry data, which is
known from well test or from manifestation analysis..

3.8 Fire and Extended reach

Fire in the wellbore may happen because the drilling fluid consist air, oxygen and other flameable substances. Bennion et. al.
(1998) has introduces a chart, it shows that more air within the drilling fluid will give a bigger possibility to trigger a fire. With
some certain pressure, fire will happen. Some underbalance drilling fluid has it’s own advantage related to fire in the wellbore.
Nitrified mud has some advantages which are: less corrosion possibility if compare with water, if oil is used as base mud then no
corrosion will happen, Less down hole fire hazard. Other hazard is that underbalance drilling fluid will give more rapid corrosion
activity on drilling equipments. Proper corrosion inhibitor must have to get bigger attention. Fire in the wellbore and rapid
corrosion tendencies have to take in bigger consideration to make sure safety first.

Extended reach drilling is a drilling process which has more than twice longer directional section than vertical section. Drilling
survey data will give information about what inclination is happening on certain depth..
4
 Dwinanto and Rachmat

4. BOTTOM HOLE PRESSURE WINDOW AREA CALCULATION

To determine a bottom hole pressure window, five criteria has to be calculated (Arliyando, 2012). These criteria are:

 Pore pressure
 Bottom hole pressure
 Minimum bottom hole pressure stability
 Maximum pump rate
 Hole cleaning assessment

Pore pressure is extracted from pressure and temperature surveys at feedzone on each well. Afterward pore pressure can be looked
at some future period of production time by, simulating it on a reservoir simulation process.

Minimum bottom hole pressure stability is extracted from borehole collapse pressure gradient process. The information is about
borehole collapse pressure gradient. Then at the feedzone depth, borehole collapse pressure will be known. Maximum pump rate is
extracted from the technology available on the market. The unit that governs the limit are gallon per minute. This upper limit
maximum pump rate will limit the maximum flow rate available.

4.1 Bottom hole pressure calculation

A bottom hole pressure is a result of the drilling mud properties and the pressure loss that occur along the conduit that a drilling
mud suffer from. The drilling mud properties such as densities will give a pressure column for a given depth. And as the drilling
mud go thru a sequence of conduit, it suffers from pressure loss. Pressure loss occur mainly on frictions.

A multiphase flow can be characterized as : bubbly, slug, churn and annular flow. Experiences gained from well control indicate
that bubbly flow dominates the aerated mud flow in drilling operations. It is a reasonable assumption that the aerated mud flow can
be treated as a homogeneos mixture of liquid, gas and solid if it is flowing in bubbly regime for the purpose of pressure calculations
(Guo et al., 1996). A homogeneos gives an idea that those three phase flow within a same velocity, thus a simpler calculation.

Lyon, et al. (2009) has a method of calculation for a homogeneous - aerated drilling method. The method for pressure calculation at
any position is a product of pressure losses from frictions and mud column pressures. This equation 8 has to be recognized with
some subtitution acting as V as velocity, f as friction factor and γmix as mix density of the drilling fluid. The result for the
subtitution is on equation 10. A step by step homogeneous - aerated drilling method iteration process is needed and can be
implemented with the flow chart given on Air and Gas Drilling Manual by Lyon, et al. (2009).

(8)

(9)

(10)

4.2 Hole cleaning assessment

Lyon et. al. (2009) said that the cleaning, lifting, and suspension capabilities of the incompressible drilling mud is generally
independent of the depth of drilling. Conversely, the cleaning and lifting capabilities of compressed gas are dependent of the depth
of drilling. Also, it must be noted that compressed gas drilling fluids have little or no suspension capabilities. Therefore, when
designing an aerated drilling fluid, the injected compressed gas should not be assumed to contribute to bottom hole cleaning, lifting,
and suspension of rock cuttings in the annulus. The additional cleaning and lifting properties of the compressed gas to the aerated
drilling fluid should be considered bonuses. On this case incompressible liquid is somewhat to rely on lifting cuttings.

Lyon et. al. (2009) explain about steps to search minimum velocity value. It begins with calculating critical concentration velocity.
The cutting concentration limit is 0.04. This value sets a maximum cutting concentration before it makes some trouble with hole
cleaning. Then assuming a terminal velocity value. On this case, turbulent terminal velocity is being counted. The terminal velocity
affects the flow type which is counted by a spesific reynolds number. The challenge is that on some terminal velocity number has to
match on a result for the terminal velocity at the end. So goal seek has to be implemented on the calculation. To know the friction
factor, a spherical constanta value has to be choosen. The perfect spherical will be = 1 and a bad spherical shape will be closer to =
0 (see Air and Gas Drilling Manual by Lyon, et al. (2009) for further information). On this case, the spherical value 0.8 is choosen.
Then friction factor will be known with a spesific reynolds and spherical value. After that terminal velocity can be counted. By
knowing terminal velocity and critical concentration velocity, then minimum velocity can be counted.

4.3 Bottom hole pressure window area

Three criteria of bottom hole pressure window can be generated based on figure 14 below. The first step is to guess at what rate an
underbalance drilling should be implemented. Most aerated drilling operations are planned with a constant flow rate of
incompressible drilling fluid and only the volumetric flow rate of the compressed gas is allowed to vary. The volumetric flow rate
of gas is usually increased as the depth is increased in order to maintain the same aerated fluid properties in the annulus column
(Lyon et. al., 2009).
5
Dwinanto and Rachmat

After calculating bottom hole pressure (BHP) at every guess gas flow rate and mud flow rate, then three graphic is made :
 Bottom hole pressure (BHP) vs Qgas graphic
 Velocity vs Qgas graphic
 Qmix vs Qgas graphic

Velocity vs Qgas graphic needs additional calculation from hole cleaning assessment and actual velocity calculation. Two graphic
which are Qmix vs Qgas graphic and velocity vs Qgas graphic are inputs for the bottom hole pressure (BHP) vs Qgas graphic.

Hole cleaning assessment will generate a new data series of Vminimum for a given Qgas. Another data series will be generated
from actual velocity calculation. This two data series will give a different gradient thus give the possibility to intersect one and
another. The intersection give a minimum ability at a certain Qgas (P) to be implemented. This Qgas (P) value then will be an input
for a new bottom hole pressure calculation. The result Qgas (P) and bottom hole pressure series then are one of the criteria for the
bottom hole pressure window.

A mud pump has a certain maximum ability to pump a mud. This maximum pump will generate one data series. Another data series
is generated by the bottom hole pressure calculation, with Qgas and Qmix data series. This two data series will give a different
gradient thus give the possibility to intersect one and another. The intersection give a maximum ability at a certain Qgas (P) to be
implemented. This Qgas (P) value then will be an input for a new bottom hole pressure calculation. The result Qgas (P) and bottom
hole pressure series then are one of the criteria for the bottom hole pressure window.

5. STUDY CASE
5.1 Field overview
The reservoir is a hypothetic model very simmilar to one of a high enthalpy geothermal field. This field has a maximum area of 20
km2. “X” geothermal field is located in a highly volcanically active. The dominant tectonic structure is a large explosion crater
approximately 1 km wide, which on the surface is observed as a circular caldera subsidence feature, and surrounded by a tuff cone
ring. This prospect area contains the large caldera lake which was produced in Plio-Pleistocene times. “X” geothermal prospect is
located in a relatively flat depression with an average elevation of 750 m a.s.l. The prediction using monte carlo corelation,
potential to gain electricity is for P50 probability to generate 46 Mwe for 25 years.

It had been drilled 8 wells so far, which consist of 2 exploration wells (X-1 and X-2), 4 production wells (X-3, X-4, X-6 and X-7)
and 2 injection wells (X-5 and X-8). From pressure and temperature gradient simmilarities three reservoir cluster can be confirmed.
Cluster 1 consist of X-1 and X-3. This cluster has a relatively good potential of permeability. Cluster 5 which consist of well X-5
and X-8 has not entered high temperature zone. But from the transmisivity data, it shows that these wells had not intersected much
fault networks. Cluster 2 is the most potential for power generation. These wells (X-2, X-4, X-6 and X-7) have enough high
temperature and good permeability to rely on. Table 4 and table 5 will give the clearest information.Three wells proved to produce
a certain amount of vapor : X-2, X-4 and X-7 with a capacity for 5 MW, 5 MW and 12 MW respectively. From a research their
Cl/B and Cl/Li geochemistry ratios indicated that X-2 and X-7 is from a same reservoir fluid.

5.2 Casing diameter feasibility

The next step is to simulate the wellbore. Wellbore simulation use Hagedorn & Brown (1965) method. The reason is that base on
Hasan & Kabir (2010) research, Hagedorn & Brown gives the biggest R-squared result if comparing with other methods. It is
expected that this method gives the least error. The simulation goes thru all of the calculation using a wellbore simulation software.
Here are the results as an WHP vs Q total graphic and WHP vs enthalpy graphic. It shows a great match for the simulation results
and the actual datas. (see figure 3 below). Productivity index result PI = 1.12 kg / s . bar for X-2 and a value PI = 2.12 kg / s . bar
for X-7. To simplify the next step it is assumed that productivity index for X-7 gives a close result with the future makeup well.

WHP vs Q Total WHP vs Q Total

200 200
Q Total (kg/s)

Q Total (kg/s)

100 150
100
0
0 20 40 60 50
WHP (bar) 0
0 20 40 60
Simulation curve Actual data WHP (bar)

Figure 3 (Left side) Wellbore simulation results for X-7;

Figure 4 (Right side) Wellbore simulation results for X-7 production casing 13 3/8 "
Figure 4 is the simulation result when X-7 well is implemented with big hole production casing (casing 13 3/8 "). Then a casing
diameter feasibility is conducted for several casing diameters. With a limit of 2.44 m/s velocity in a steel pipe, then a decision for a
good casing diameter implementation design can be calculated. Table 6, table 7 and table 8 gives a best calculation picture.

6
 Dwinanto and Rachmat

5.3 Underbalance screening process

Figure 5 describes the flow of this decision. The black bold shapes will give the path towards underbalance decision. Based on
three assessment, underbalance method has the opportunity to reduce reactive clay problems which will induce some stability
issues. On a overbalance method basis, a swelling may exist from montmorillonite scattered on the clay formation. Although with
some lack of rock strength data for weak formation, these stability issues have to be added some more calculation with more
wellbore stability analysis. But furthermore, if considering that it surpasses stable rock requirement, then other assessment need to
be realize. Subsequently lost circulation assessment on table 3 has showed that cluster 2 and cluster 5 have some lost possibilities.
At the other part cluster 1 has severe lost possibilties. The last assessment is for reservoir damage. Reactive clay results are 28%
possibility of swelling problem and 68% possibility of fine migration possibility on the reservoir area. Thus the conclusion is the
reservoir is prone for reservoir damage. From all three assessment underbalance method may have the opportunity to reduce
problem that may exist on overbalance method regardless about the lack of wellbore analysis.

Table 3 Lost circulation calculation

Cluster Cluster 2 Cluster 5 Cluster 1
Fracture Index 0.75 3 2
Vug Index 3.75 3.75 3.75
Porosity 0.15 0.15 0.15
Transmisivity (md) 8.87 0.9 4.63
Lost Circulation Index 0.04 0.02 0.05
Lost Circulation Possibilities some LC some LC severe LC

From figure 6 describe the vertical wellbore drilling fluid decisions. Based on the well’s pressure and temperature chart, it had been
justify that “x” field is a liquid dominated reservoir. On liquid dominated reservoir, liquid is the most available fluid. Casing
feasibility study shows that it needs a big hole for a minimum casing production 13 3/8 ". The next criteria is water influx. It has
been stated that reservoir is a liquid dominated type, thus the criteria will be definetly has water influx. The last criteria is about gas
existence. The only well that give a clue about 2 phase dominated well is X-7. Another clue about gas existence is two fumarol
which is located at the West side of X-1 and at the far East side of X-7. So the probability to found another two phase well is still
big although with a slight fraction in fluid properties. So it is assumed that the gas existencies is not significant. At the end stiff
foam or aerated mud is a final choice for vertical well with underbalance drilling method (If a horizontal wellbore will be
conducted, then figure 7 has to be implemented). The black bold shapes will give the path towards underbalance decision.

5.4 Hole cleaning and bottom hole pressure calculation

For hole cleaning assessment, a full scale calculation had been made with microsoft excel. The iteration uses goal seek feature on
microsoft excel. This microsoft excel sheet may help other engineer to search minimum velocity for certain parameters. The result
of the minimum velocity is 1.88 ft/s.

A full scale bottom hole pressure calculation had been made with microsoft excel. The iteration uses goal seek feature on microsoft
excel. A few calculation are made under macro-excel. This microsoft excel sheet may help other engineer to search pressure at
every section and actual velocity at every section at every inner annulus area differentiations. Calculation is implemented on liquid
flow rate (Q liquid) on 0.1 ft3/s, 0.15 ft3/s, 0.2 ft3/s, 0.3 ft3/s and 0.4 ft3/s.

5.5 Bottom hole pressure window determination

From the X-7 well feedzone, it is concluded that the pore pressure is 1226.05 psi. With a reservoir simulation for 5 years production
time a pore pressure is concluded for 1153.53 psi. This pore pressure will give an upper limit for a window area.

For this study two pumps are included. Neyrfor Weir Motor Max pump with a max capacity for 350 gpm and White Star Quatro
2200 hp with a max capacity for 1597 gpm. These two pump will give a max limit on the bottom hole pressure window.

This study results are in figure 12, figure 13 and figure 15. Neyrfor Weir Motor Max pump did not give a satisfaction window area.
White Star Quatro 2200 hp gives a little window area to be implemented. The results give 2 flow rate alternatives settings:

 Qliquid at 0.15 ft3/s and a range of Qgasat 3.10 ft3/s – 3.25 ft3/s
 Qliquid at 0.10ft3/s and a range of Qgasat 2.10 ft3/s – 3.20 ft3/s

On those two alternatif, care must be taken. Because the minimum bottom hole pressure collapse has not to be known yet. And
alternatif one only gives a limit of underbalance diferential from pore pressure @5 years for around 30-50 psi. This is not an ideal
condition since a bottom hole pressure fluctuation on drilling connections (Saponja, 1998) needs a minimum safety factor for
connection condition for about 350 psi. If this alternatif is applied then when pipe connection, it will absolutely become
overbalance. This will result mud to seep into the vugs and fracture networks. Which will cause problem like formation damage,
pipe sticking and lost of valuable fluids.

At the second alternatif (alternatif number two) is more peferable. Considering the bottom hole pressure fluctuation on drilling
connections (Saponja, 1998) addtional factor, then a range of Qliquid at 0.10ft 3/s and a range of Qgas around 2.90 ft3/s – 3.20 ft3/s
should be implemented.

6. CONCLUSION
The conclusion is that this hypothetical field has a great chance to implement aerated underbalance drilling method. With a
minimum hole cleaning assessment velocity of 1.88 ft/s, two alternatif had been produced. The best alternative is to implement a
range of Qliquid at 0.10ft3/s and a range of Qgas around 2.90 ft 3/s – 3.20 ft3/s should be implemented.
7
Dwinanto and Rachmat

7. DISCUSSION
Wellbore stability quantitative analysis for borehole collapse pressure has not been conducted, but it absolutely need to be
conducted. This analysis will help this bottom hole pressure window area research to be more reliable. An economic analysis have
not been conducted. Future research can be conducted on this same reservoir for stiff foam mud. These subjects can be the future
researches.

NOMENCLATURE
Q : flow rate (m3/s)
ID : casing inside diameter (m)
V : fluid velocity (m/s)
P : pressure drop (bara)
ρ : fluid spesific weight (kg/m3)
g : gravity acceleration (9.81 m/s2); (32.2 ft/sec2)
v : fluid velocity (m/s)
z : well depth (m)
d : pipe diameter (m)
f : friction factor (dimensionless)
: lost circulation index (%)
k : absolute permeability (md)
: porosity (%)
: fracture index (Scale 1-10)
: vugs index (Scale 1-10)
SDSI : sharif’s differential pipe sticking index
CC : cutting concentration (%)
SOL : solid percentage in drilling fluid (%)
EMW : effective mud weight (lb/gal)
Dia : hole diameter (ft)
ROP : rate of penetration (ft/hr)
: differential pipe sticking probability (%)
: effective porosity value (%)
: total Fine-migration Index of the particular clay
: volume fraction of the particular clay
: fine-migration index of the particular clay.
: fine-migration factor of the particular clay.
: swelling potential
: total swelling index of the particular clay
: distribution correction factor of the particular clay for swelling.
: fine-migration potential
: distribution correction factor of the particular clay for fine-migration.
: minimum velocity (ft/s)
: critical concentration velocity (ft/s)
: rate of penetration (ft/s)
: reynolds number (dimensionless)
ν : kinematic viscosity (ft2/s)
: mix fluid specific weight (lb/ ft3)
: fluid friction factor (dimensionless)
V : fluid actual velocity (ft/s)
: wellbore diameter (ft)
: outside pipe diameter (ft)
: depth (ft)
: bottom hole pressure (lb / ft2)
: upper section pressure (lb / ft2)
̇ : total weight rate (lb /s)
: atmosfer pressure (lb / ft2)
: average temperature; (oR = 459.67 + oF)
: atmosfer temperature; (oR = 459.67 + oF)
3
: gas flow rate (ft /s)
: mud flow rate (ft3/s)
: resistance coeficient (0.2)
: blind tee resistance coeficient (30)
A : fluid cross sectional flow area (ft2)
: engineering gas constant; (53.36 ft.lb/lb.oR)
: liquid dynamic / absolute viscosity (lb.s/ ft2)`
: reynolds number (dimensionless)
WHP : well head pressure
: total weight rate of fluid (kg/s)
TLC : total lost circulation
PLC : partial lost circulation
8
 Dwinanto and Rachmat

REFERENCES
Al-Ajmi. (2003). Optimum Selection of Underbalanced Techniques. Texas: Texas A&M University.
Arliyando, L. (2012). Feasibility study of Underbalanced Drilling Using Coiled Tubing: A Case Study Field : "A", Indonesia.
Bandung: ITB.
Bennion, D. B. (1998). Underbalanced Drilling And Completion Operations to Minimize Formation Damage-Reservoir Screening
Criteria For Optimum Application. Journal of Canadian Petroleum Technology 37.9.
Fuh, G.-F., Whitfill, D. L., & Schuh., P. R. (1988). Use of borehole stability analysis for successful drilling of high-angle hole.
SPE/IADC Drilling Conference. SPE.
Guo, B., Hareland, G., & Rajtar, J. (1996). Computer simulation predicts unfavorable mud rate and optimum air injection rate for
aerated mud drilling. SPE Drilling & Completion 11.2, 61-66.
Hagedorn, A. a. (1965). Experimental study of pressure gradients occurring during continuous two-phase flow in small-diameter
vertical conduits. Journal of Petroleum Technology 17.4, 475-484.
Hasan, A. R. (2010). Modeling Two-Phase Fluid and Heat Flows in Geothermal Wells. Journal of Petroleum Science and
Engineering 71.1, 77-86.
Hole, H. (2006). Aerated fluids for drilling of geothermal wells. United Nation University.
Lyle, O. (1947). The efficient use of steam. HM Stationery Office.
Lyons, W. C., Guo, B., Graham, R. L., & Hawley, G. D. (2009). Air and Gas Drilling Manual: Applications for Oil and Gas
Recovery Wells and Geothermal Fluids Recovery Wells. Burlington: Elsevier.
Mosch, S., & Siegesmund, S. (2007). Petrophysical and technical properties of dimensional stones: a statistical approach.
Zeitschrift der Deutschen Gesellschaft für Geowissenschaften 158.4 , 821-868.
Saponja, J. (1998). Challenges with jointed pipe underbalanced operations. SPE drilling & completion 13.2, 121-128.
Sharif, Q. (1997). A Case Study of Stuck Pipe Problems and Development of Statistical Models to Predict the Probability of Ge
tting Stuck and IF Stuck, The Probability of Getting Free. Texas: Texas A&M University.
Vitthal, S., Gupta, A., & Sharma, M. (1989). A Ruled-Based System for Estimating Clay Distribution, Morphology, and Formation
Damage in Reservoir Rocks. SPEFE .

APPENDIX
Start
Start ( Vertical bor )

No
Small
No Hole
Stable Rock
No UBD Anticipated Yes
No No
Yes Gas
Water
Influx
No No
Lost Yes Yes
Stuck Pipe
Circulation

Yes Yes Yes

Yes Gas
No H2S
Hard
Drilling No
No
Yes Nitrogen or Stiff Foam or
Nitrified Mud Aerated Mud

No UBD
Fire
No No
Cost Reservoir
No UBD Benefit Damage No Yes

Yes Yes

UBD UBD Dry Air or Mist or

Aerated Mud Nitrified Mud

Figure 5 (Left side) Underbalance decision flow chart (Al-ajmi, 2003);

Figure 6 (Right side) Drilling fluid decision on vertical wellbore flow chart (Al-ajmi, 2003)

Figure 7 Drilling fluid decision on horizontal wellbore flow chart (Al-ajmi, 2003)
9
Dwinanto and Rachmat

Density 1.2 g/cm3 1.57 g/cm3 1.93 g/cm3 2.3 g/cm3 2.67 g/cm3 3.03 g/cm3 3.4 g/cm3 Legend

cluster2

Density class 0 % 10 % 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % 100 %

(0 % - 50 %) Low Density rocks area (50 % - 100 %) High Density rocks area

Figure 8 Rock density classification; Andesit / basalt rock on cluster 2 is classified as high density rock

Porosity 0 0.08 0.15 0.23 0.31 0.39 0.46 Legend

cluster2

Porosity class0 % 10 % 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % 100 %

(0 % - 50 %) Low Porosity rocks area (50 % - 100 %) High Porosity rocks area

Figure 9 Rock porosity classification; Andesit / basalt rock on cluster 2 is classified as low porosity rock

LC event 0 1 5 9 13 17 21 25 30 Legend

Cluster 2

Cluster 1

Cluster 5

LC index 0 1 2 3 4 5 6 7 8 9 10 Y - wells (Sumatra)

Figure 10 Lost circulation classification; Cluster 2 rock is classified as low fractured rock thus low permeability

Figure 11 Bulk density, effective porosity and water absorption on each rock types. Based on 2335 stones (Mosch and
Siegesmund, 2007)
Table 4 Reservoir characteristic cluster-1 and cluster-5
Cluster - 1 Cluster - 5
Well X-1 X-3 X-5 X-8
Reservoir type Water Water Water Water
Reservoir
225-250oC 225-300oC 225oC 225oC
temperature
Transmisivitas
n/a 4.63 0.9 0.23
(kh), Darcy.m
Skin n/a -0.04 n/a n/a
Well depth -1142 mrsl -1122 mrsl -1050 mrsl -1081 mrsl
Total loss
No TLC -1050 mrsl No TLC No TLC
circulation (TLC)
-427 mrsl; -593
134 mrsl; -7 mrsl; -496 mrsl; -785 mrsl; -90 mrsl to -171 mrsl;
Partial loss mrsl;
circulation (PLC) -26 mrsl; -35 mrsl; -879 mrsl; -894 mrsl; -219 mrsl to -372 mrsl;
-922 mrsl
-48 mrsl; -694 mrsl; -945 mrsl; -1020 mrsl; -418 mrsl to -1014 mrsl;
10
 Dwinanto and Rachmat

Table 5 Reservoir characteristic cluster-2

Cluster - 2
Well X-2 X-4 X-6 X-7
Reservoir type Water Water Water two phase
Reservoir
260oC 225-255oC 257oC 250-270oC
temperature
Transmisivitas (kh),
n/a 7.12 n/a 8.87
Darcy.m
Skin n/a -1.67 n/a n/a
Well depth -800 mrsl -1000 mrsl -1019 mrsl -714 mrsl
Total loss -280 mrsl and
-400 mrsl -400 mrsl -200 mrsl
circulation (TLC) -550 mrsl
Partial loss -144 mrsl; -360
-389 mrsl No PLC -194 mrsl
circulation (PLC) mrsl;

Table 6 Casing diameter feasibility study (13 3/8 ")

Casing Production 13 3/8 "

Well Weight Saturated
Flow Velocity
Head Flow Water Spesific Casing properties Result
Rate of liquid
Pressure Rate Volume
ID A
bar kg/s m3/kg m3/s A (m2) m/s
(inch) (inch2)
16 140 0.00 0.17 12.25 117.86 0.08 2.24 Approved

Table 7 Casing diameter feasibility study (10 3/4 ")

Casing Production 10 3/4 "

Well Weight Saturated
Flow Velocity
Head Flow Water Spesific Casing properties Result
Rate of liquid
Pressure Rate Volume
ID A
(bar) (kg/s) V (m3/kg) m3/s A (m2) m/s
(inch) (inch2)
16 140 0.00 0.17 9.56 71.78 0.05 3.68 Bad Design

Table 8 Casing diameter feasibility study (7 ")

Casing Production 7 "

Well Weight Saturated
Flow Velocity
Head Flow Water Spesific Casing properties Result
Rate of liquid
Pressure Rate Volume
ID A
(bar) (kg/s) V (m3/kg) m3/s A (m2) m/s
(inch) (inch2)
16 105 0.00 0.13 5.92 27.53 0.02 7.20 Bad Design

Figure 12 Velocity (V) vs Gas flow rate (Q gas) graphic

11
Dwinanto and Rachmat

Figure 13 Mix flow rate (Q mix) vs Gas flow rate (Q gas) graphic
Start

Pick a range of Qgas on a spesific

Qliquid

Calculate Bottom Hole Pressure at Calculate V minimum with hole

every Qgas for a given Qliquid cleaning assessment at every Qgas

Generate graphic : BHP vs Qgas with Calculate actual fluid velocity at the
series from BHP calculation widest area of all sections

Generate graphic : Qmix vs Qgas.

Implement 2 type of series from the Generate Velocity vs Qgas graphic
max pump capacity and the BHP with 2 types of series : Vactual and
Vmin from above
calculation

Identify at what Qgas(P) the max

Identify at what Qgas(P) the Vactual
pump series intersect the BHP
series intersect the Vmin series
calculation series

On Qgas(P), generate the BHP On Qgas(P), generate the BHP

calcuculation. Then implement the calcuculation. Then implement the
series of BHP at Qgas(P) at BHP vs series of BHP at Qgas(P) at BHP vs
Qgas graphic Qgas graphic

Combine with these series on

graphic BHP vs Qgas

Identify Bottom Hole Pressure

window at graphic BHP vs Qgas

End

Figure 14 Bottom hole pressure window area determination

Figure 15 Bottom hole pressure (BHP) vs Gas flow rate (Q gas) graphic
12