SPE-184846-MS

Cracking the Volcanic Rocks in India Substantial Benefits From Continuous

Improvements Over 11 Years and 100 Fracturing Treatments

Shobhit Tiwari, Raymond Joseph Tibbles, Shashank Pathak, Saurabh Anand, Yudho Agustinus, Punj Siddharth,
Rajat Goyal, Vishal Ranjan, Pranay Shrivastava, Hindul Bharadwaj, Pranay Shankar, and Leste Aihevba, Cairn
India Ltd.

Copyright 2017, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Hydraulic Fracturing Technology Conference and Exhibition held in The Woodlands, Texas, USA, 24-26 January
2017.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Objective: This paper summarizes key engineering discoveries and technical findings observed during
the staged development of a volcanic reservoir. Through out the development, 200 hydraulic fracturing
diagnostic injection tests and 168 hydraulic fracturing treatments were performed. This program was
conducted in one of the few commercially viable thick and laminated volcanic gas reservoirs in the world
and were staggered in 5 separate campaigns over an 11-year period.
Method, Procedure, Process: Due to the low permeability of this gas reservoir, hydraulic fracturing was
necessary for sustained economic productivity. As this massive laminated reservoir contained between 15
to 40 vertically separated pay sections, a key design consideration was to connect as much pay as possible
with the least number of fracturing stages.
Although a conventional plug and perforation frac technique gives full assurance of optimal fractures for
every bit of pay, the completion cost would undermine the project's economics. Therefore, the limited entry
technique was selected. The uncertainties and risks were evaluated to maximize the probability of success.
Both methodologies were applied in successive campaigns. Staggering the project into 5 successive
campaigns enabled adequate time to evaluate results, acquire data and execute. The results from these
learnings are summarized in this paper.
Results, Observations, Conclusions: Over 60 DFITs (Diagnostic Fracture Injection Tests), ~90 SRTs and
~50 Mini-Fracs have been conducted. In addition to conventional fracture diagnostics tests, other techniques
were applied with successful implementation. One such example was the utilization of multiple step rate
tests within the same frac stage to evaluate limited entry efficiency. As a result of the test data, the number
of clusters per frac was increased from 3 to 6, increasing the net pay coverage by about 65%.
Another achievement was the reduction of the uncertainty in tubing friction and the evaluation of tubing
friction increase due to the addition of proppant. This resulted in a cost effective method of reducing
uncertainties in calculated BHPs, thus improving the overall understanding of fracture geometry. This paper
also demonstrates that the integration of all of the collected diagnostic data, temperature surveys, frac
2 SPE-184846-MS

simulation and geo-mechanic calibration resulted in increased contribution from more zones which was
verified with production logs.
This enhanced reservoir understanding greatly helped to save operational time and reduce cost.
Completion improvements have resulted in an 80% increase in productivity and a 20% increase in EUR.
Screen out rates have dropped from 33% to 5% between the initial and the most recent campaign.
Novel/Additive Information: A holistic workflow for conducting diagnostic injection tests in volcanic
pays.
Detailed analysis of limited entry controlled hydraulic fracturing and its efficiency.
Representative case histories including, DFITS, Step rate tests, Mini Fracs, Temperature surveys and
production logs to back up the production results.

Introduction
The Raageshwari Deep Gas (RDG) Field in the southern Barmer basin, India, is a thick (~700 m gross),
low permeability (0.01-1 md) gas-condensate reservoir with excellent gas quality of approximately 80%
methane, low CO2 and no H2S. The gas has net calorific value of about 1070 BTU/SCF and the condensate
gravity is ~ 56 API. The field was discovered in 2003 and an initial field development plan (FDP), was
approved in 2006, focusing on internal fuel consumption for oilfields located in the northern part of the
development block. A comprehensive re-evaluation of the resource base was carried out based on four years
of production history and additional data acquired and a revised FDP for commercial gas sales was approved
in 2014. Multi-stage hydraulic fracturing is being implemented across the field to optimally produce the
gas reserves in these tight volcanic reservoirs.

Reservoir Background
The RDG field is situated at the northern end of Central Basin High (CBH) a 40km-long composite feature
of elevated N-S-oriented fault terraces in the southern Barmer basin (see Figure 1). The targeted pay zones
consist of a poorly sorted clastic (Fatehgarh) interval on top of a stacked volcanic succession of low net-to-
gross lava flow cycles of Basalt and Felsic units (see Figure 2). Onlap of Fatehgarh fluvial and lacustrine
sediments were deposited onto the volcanic high during syn-rift faulting created by dominant -N-S faults.
The average depth of the reservoir ranges from 2,500 m TVDSS to below 3,500 m TVDSS. The reservoir
is a low NTG system with a high degree of heterogeneity. Average porosity ranges from 8% in the volcanic
to 12% in the clastic reservoirs, with a relatively high average connate water saturation of 50%.
SPE-184846-MS 3

Figure 1RDG Map

Figure 2RDG Structure

Given the volcanic nature of the field, conventional sand-shale and carbonate-shale workflows are not
applicable to estimate various petrophysical parameters. NMR being a lithology independent log was used
along with other basic logs (GR, neutron, density and resistivity) for facies identification using a proprietary
workflow. Geomechanical properties estimated from sonic and image logs were integrated with analog
data from other volcanic fields to build reasonable 1D geomechanical models for all wells. The integrated
petrophysical model coupled with the 1D geomechanical model for each well were used to identify the
optimum completion intervals.
Figure 3 shows an example of the integrated log. Details of the workflow can be found in the paper
presented by Mund et al at the 2016 Petrotech conference. (Bineet Mund, 2016)
4 SPE-184846-MS

Figure 3Frac Planning Log for Well W

RDG field started production in March 2010 but targeting only internal fuel gas demand from northern
oil fields. In April 2013, commercial gas sales started which allowed extensive data acquisition and well
surveillance in several wells. Currently, 30 wells have been completed with 168 hydraulic fractures. A little
over 80% of these fractures were in the volcanics (Basalt and Felsic).

Completion Design
To date, all of the Raageshwari gas well completions were 3 " monobore vertical wells with 9.2 ppf P110
tubing to surface. This design was dictated by the worst tubing load observable during the well life cycle
which would occur during the fracturing operation in a screenout to surface incident (Jha, 2010). All other
loads (frac diagnostics, routine frac pumping, and well production) were found to apply considerably lower
loads on the completion. Figure 4 shows the details of the current well design.

Figure 4RDG Well Completion - Old Design

SPE-184846-MS 5

Maximum surface pressure limitations were one of the constraints on injection rate. Due to the mix water
quality and the corresponding frac fluid properties (discussed later), tubing friction in the last campaign
was higher than expected which resulted in lower maximum achievable pump rates for the lower and
intermediate zones. Since water quality is expected to remain the same, the only way to reduce friction and
increase maximum rate is to increase the tubing size. Increasing the pump rate allows more clusters to be
treated in a single pumping operation which will reduce the total number of stages and therefore the cost of
the well. The comparison of tubing friction in Figure 5 shows that if we assume that the limitations for total
tubing friction stay the same which allowed for 25 bpm in the 3 inch completion, injection rates between
35 and 40 bpm can be achieved by either partially or completely replacing the 3 inch pipe with 4 inch
pipe. Another constraint in the well design is the required gas velocities for liquid unloading (especially
in the deeper zones).

Figure 5Comparison of RDG frac fluid friction in 3.5 and 4.5 tubing

As part of the completion design process for the next phase of development, well modelling was done
using a transient-state simulation software to quantify the critical rates (CRs) required to prevent water
loading during well cleanup in 3-1/2" and 4-1/2" monobore completions. The model assumed that a) all frac
stages have an equal PI, and b) the reservoir fluid has a WGR of 100 stb/MMscf (this high WGR was used
to mimic the high WGR seen during frac flowback).
The model was calibrated using the available flowing gradient survey data. It was then run to predict
the CRs at different FTHPs. The CR calculated for 3-1/2" liner was 560 Mscfd and for 4-1/2" liner was
900 Mscfd. The CRs calculated by the model were validated using PLT data. The cumulative rate from the
frac stages above which no static water column was observed in the PLT was taken to be the CR. The CRs
measured in such a way from the PLTs matched well with the theoretical CRs (see Table 1). Since the wells
in RDG have a 3-1/2" monobore completion the critical rates for 4-1/2" could not be verified. By combining
the CR results with the requirements for fracturing, the optimum design for future wells was found to be a
3 inch liner with 4 inch tubing (see schematic in Figure 6).

Table 1Critical Rate Measurements

Well PLT CR Simulated CR (Mscfd) % Diff

(Mscfd)

N 487 560 13%

T 531 560 5%

X 638 560 14%

Y 648 560 16%

6 SPE-184846-MS

Figure 6RDG Well Completion - New Design

Lessons from Early Frac Campaigns (2006 to 2010)

The objective of the 3 initial fracturing campaigns was to provide energy for Cairn India's Rajasthan oil
fields/processing terminal. To provide the required energy, 39 frac stages were pumped in 11 wells over
a 4 year period from 2006 (Shaoul, 2007) to 2010 (Vermani, 2010). Figure 7 shows a summary of these
campaigns. Note that in 2010 the number of stages was significantly increased to maximize net coverage.

Figure 7Initial Frac Campaigns

SPE-184846-MS 7

The frac models used at that time predicted frac heights from 50 to 60 meters. Since the total gross height
was between 250 and 300 m, 4 to 5 frac stages per well were deemed suitable to cover the entire interval.
Gas shows were used as the primary indicator for selecting the perforation points.
The fracture design also included the following:

High fracture conductivity due to non-Darcy effects associated with multiphase flow (wet gas
reservoir).
Max proppant concentration of 8 ppa which resulted with 2 lbs/ft2 targeting in the fracture.

The average job size was 250K lbs of proppant.

Fracture heights of 50 m with half lengths of about 200 m.

A borate cross-linked guar system with a polymer loading of 35 lb/1000 gal.

With the 3-1/2" mono-bore completion, successive frac stages were pumped utilizing conventional plug
and perforation technology. Both E-line based explosive perforating and coiled tubing sand jet perforations
(Arora, et al., 2011) were used and evaluated.
Despite the findings, many challenges still existed in fracturing the RDG wells. Although the screen out
rate dropped from 33% to 18% over the course of the three campaigns, this was still much higher than the
industry average of about 5%. Operational issues related to proppant admittance, premature screen outs,
and poor injectivity continued to plague the fracturing treatments.
Questions about the actual frac heights and corresponding fracture geometry achieved also remained
unanswered.

Technical Risks and Workflow to Address Them

In 2013 the primary objective of the field production changed from internal consumption to commercial
sales and further development was required. Because of this, the previous treatments were reevaluated to
identify risks (challenges and uncertainties) associated with the project.
As this massive laminated reservoir contained between 15 and 40 vertically separated pay sections,
accurately predicting fracture height and the associated pay coverage was essential. Temperature surveys
were chosen as a means for measuring fracture height because they were the most cost effective means
available. Temperature logs have long been used to infer fracture heights in vertical wells (Dobkins,
1981) (Davis, 1997). Since proppant settling/fill would inhibit successful temperature logging below the
perforations, the surveys were run after both SRT's and minifracs as well as the propped fracs.
The initial temperature survey results seen at the beginning of the 4th frac campaign indicated the fracture
was much more contained than previously thought. This meant that only 40% to 50% of the productive
intervals would be covered with 5 conventional frac stages (see Figure 8).
8 SPE-184846-MS

Figure 8Pre and post evaluation fracture height comparison

In addition, improved petro-physical understanding increased the number of available pay sections. The
challenge was how to obtain the 70 to 80% pay coverage without significantly increasing the number of frac
stages. Pumping a separate frac for each pay interval was uneconomical although it would have ensured the
best coverage of all layers. As a result, limited entry was selected as a reasonable alternative to connect as
many zones as possible with the least number of stages.
The workflow shown in Figure 9 was generated to provide a framework for improving fracturing results.
The first step was to calibrate the log stress in pays using closure pressured inferred from extended DFIT's
from the pay sections. Once the stresses were confirmed, the variation between the pays was statistically
evaluated. The observed average stress variation between the pays was less than 400 psi. This was the basis
for initial limited entry perforation design.

Figure 9Treatment design and evaluation workflow

SPE-184846-MS 9

Given a 400 psi maximum pay stress difference, 600 psi of perforation friction was deemed adequate
to ensure complete fracturing of each perforation interval. Modelling indicated that twenty-four 0.29 inch
diameter perforations will provide 600 psi friction at the expected injection rate of 30 bpm.
Since this was a new technique, the base design specified a maximum of 3 perforation clusters. Multiple
runs of the frac simulator were used to determine how to distribute the 24 holes between the three clusters.
In general, one or two smaller pays were combined with a larger pay to maximize the chances of pumping
the full treatment.
After perforating per the design, a pre-frac SRT was conducted. (See an example SRT from well Y in
Figure 10). The objective of the SRT was to determine the perforation pressure drop and ensure that it
was at or above the minimum 600 psi. The perforation pressure drop was evaluated using a step down
(SD) test at the end of the SRT. The pressure drops at different rates were compared with estimated values
to flag any unusual cases. After approximately 30 SD tests were analyzed, it was clear that the observed
perforation friction was much higher than the predicted pressure. This led to an estimate of between 50 and
75% perforation efficiency (% of perforations actually taking fluid).

Figure 10Example SRT from well Y

Post SRT temperature surveys were used to confirm that all the clusters were taking fluid. Figure 11
shows the temperature survey taken after the SRT in well Y. Note that each of the perforation clusters exhibit
cooldown which is an indication that they are all taking fluid.
10 SPE-184846-MS

Figure 11Post SRT temperature survey for well Y

To get a better understanding of the actual perforation friction, a good estimate of tubing friction was
essential. Tubing friction numbers for the various fluids (brine, linear gel, crosslinked fluid) used in the last
campaign were obtained using a series of injection tests which were run with bottom hole pressure gauges
in the well. The quantification of the perforation efficiency along with the temperature surveys increased the
confidence in limited entry and allowed us to increase the number of clusters treated. It was also estimated
that the total number of perforations could be increased from 24 to 30 while maintaining enough pressure
drop for diversion. The verification of the perforation efficiency and the fact that all clusters were taking
fluid provided the confidence to increase the number of clusters from 3 to 4.

Frac Design
Methodology
This project followed a workflow that has been described by many authors over the years. First of all, a 1d
Mechanical Earth Model (MEM) was generated for each well using the available logs and petrophysical
knowledge (see reservoir section). These MEM's were then used in the hydraulic fracturing model for the
pre-design work. Once in the field, the models were calibrated with measured data from the various injection
tests (DFIT, SRT, minifrac) and production logs (principally temperature logs).
During the initial phases of the project, calculated stress difference between the pay sections and the
barriers was quite low. Simulations therefore resulted in extensive fracture height growth which covered
multiple pay sections. In these early wells, the perforated intervals were generally larger (anywhere from 1
to 5 m) and suffered from more near-wellbore/complexity issues. Since those initial campaigns, better data
and analysis have shown that the stress contrast between the pay zones and the barriers is much higher than
previously thought. Therefore design changes were required to ensure adequate net pay coverage because
the fractures were better contained.
The greatest challenge in a low net to gross laminated reservoir is connecting all the available net reservoir
storage (h) and flow capacity (kh) using the minimum number of frac stages. The 10 or more conventional
single interval frac stages required to effectively cover the available kh (md-ft) was uneconomic.
SPE-184846-MS 11

An alternative to single perforation fracturing was tested in phase 2 of the development campaign as
stages were combined using limited entry (Lagrone & Rasmussen, 1963) fracturing. Limited entry uses the
pressure drop across the perforation to ensure fluid flow in all the intervals. The required rate per interval
is determined by the individual zone properties and stress profile. The distribution of the total available
rate then is controlled by the number and size of the perforations in each interval. For RDG, the required
pressure drop for diversion was generally in the range of 700 to 1000 psi.
This pressure drop can be calculated using equation (1) (Schumberger Oilfield Services i-handbook,
version 1,0.4,5):

(1)

Where q= flow rate, = fluid density, Cd= discharge coefficient (0.65 can be used for the initial
perforations if no other data is known), dp = perforation diameter, and n= number of active perforations
(perforations open and taking fluid). By assuming 1 perforation, Equation (1) can be rewritten to calculate
the required rate per perforation for the desired pressure drop as in equation (2).

(2)

Using a perforation diameter of 0.29 inches, a discharge coefficient of 0.85 and fluid density of 8.5 ppg,
equation (2) can be solved at various rates to generate Figure 12.

Figure 12Perforation Friction Chart

Completion and frac fluid constraints in the last fracturing campaign limited the maximum pump rate
to between 20 and 30 bpm (depending on the zone depth and insitu stress). Thus the maximum number of
effective perforations is between 16 and 30 using Figure 12 and assuming dP from 700 to 1000 psi. The
observed perforation efficiency was roughly 50%, which meant that between 26 and 48 holes could be shot
depending on the injection rate. For the first part of the campaign the number of holes were limited to 24.
This was later increased to 30 after diagnostic testing, temperature surveys and post frac pressure matching
confirmed good diversion. Therefore, the workflow consisted of taking the available number of perforations
and distributing them over 2 to 6 separate zones. In the first pass, the perforations were equally distributed.
Subsequently, the perforations were redistributed to obtain the desired fracture dimensions for each zone.
In some cases, separate individual fractures were generated while in other cases the fractures coalesced.
Figure 13 shows two examples of the two simulation extremes.
12 SPE-184846-MS

Figure 13Frac Simulation With 5 Clusters from Well T and 3 Clusters from Well S

The fracturing fluid used in the treatments was a 35 (lb/1000) borate crosslinked guar. In the last
campaign, the breaker was designed to break the fluid after the fracture closure to minimize proppant
settling. Since the time to fracture closure could be as long as 200 min, the fluid had to remain stable for
at least this long and break afterwards. In order to achieve this, encapsulated breaker was incorporated into
the system. Figure 14 shows an example of a typical RDG stability test showing working time (viscosity
greater than 100 cP at 100 sec-1) more than 200 min.

Figure 14Example fluid stability test with breaker

The main proppants used were 20/40 and 16/30 mesh ISP. A typical design schedule is showns in Table 2.

Table 2Example pump schedule

Stage Fluid Type Proppant Type Prop Conc. Prop Mass Cum Prop Stage
Mass Volume

(lbs/gal) (lbs) (lbs) (bbl)

1 Cooling Pad 0.0 - - 80

2 35# HT Borate 0.0 - - 800

5 35# HT Borate 20/40 ISP 0.5 2,678 2,678 130

Continued on next page

SPE-184846-MS 13

Stage Fluid Type Proppant Type Prop Conc. Prop Mass Cum Prop Stage
Mass Volume

(lbs/gal) (lbs) (lbs) (bbl)

6 35# HT Borate 20/40 ISP 1.0 5,257 7,935 130

7 35# HT Borate 20/40 ISP 2.0 10,137 18,073 130

8 35# HT Borate 20/40 ISP 3.0 15,809 33,881 140

9 35# HT Borate 20/40 ISP 4.0 21,829 55,710 150

10 35# HT Borate 20/40 ISP 5.0 29,923 85,633 170

11 35# HT Borate 20/40 ISP 6.0 36,828 122,462 180

12 35# HT Borate 16/30 ISP 6.0 30,923 153,385 150

13 35# HT Borate 16/30 ISP 7.0 35,015 188,401 150

14 35# HT Borate 16/30 ISP 7.5 61,600 250,000 250

Wellsite Operations
Hydraulic fracturing operations and well flowback have several challenges in Rajasthan. Being an arid
region of India, a continuous supply of water is problematic at best. In addition, the oilfield infrastructure is
much smaller than typically seen in North America with few suppliers and a dependence on small suppliers
for periphery services like water hauling. Due to these types of issues, the first campaigns suffered from
significant delays to operations leading to cost over-runs. The key issues from the previous campaigns were
evaluated and various plans put in place to ensure smoother future operations.
With the addition of these operational changes, the RDG 15 well program set new operational planning
and execution benchmarks. The number of fracs per month increased by over 400% while the cost per frac
was cut in half. The summary of the main operational challenges and how they were addressed are listed
below.
1. Rig-up and Simultaneous Operations (SIMOPS): RDG development well pads have from 8 to 12
wells per pad, distributed in two rows of cellar on either side of the production header. The well pads
cover an area of about 30,000 m2 (200 150m), of which approximately 4800 m2 is available for
rigging up of the frac spread and associated equipment. This is a relatively small area for spotting the
key equipment, including the frac spread, coil tubing unit, E-line unit and surface well test spread.
Previous campaigns suffered significant delays due to the lack of mobility around the location to
perform SIMOPS. For the RDG 15 well campaign, the layout for rig-up of the frac and associated
services was carefully planned to optimize the available area and to ensure maximum SIMOPS with
minimum downtime. Associated services such a coil tubing unit (needed for wellbore cleanout), E-
line unit (perforating, bridge plug runs, and temperature surveys) and surface well test unit were
strategically placed to ensure the ability to work and move around independently without affecting
the ongoing operations of other units.
2. Dual-well Rig up: In the previous campaigns, only one wellbore was worked on at a time, resulting
in significant time lost due to delays in bridge plug running or the need for a wellbore cleanout due to
excessive sand fill or screen-out. However, in the last campaign, two wells were simultaneously lined
up to the high pressure (HP) manifold and 10 kpsi frac trees and frac heads were installed on both
wells. In the event of a wellbore screen-out or other delay on one well, the operation could be shifted
to the second well thus eliminating equipment standby / idle time. This proved to be a very critical
decision on the few occasions of issues in setting the bridge plug or excessive sand fill requiring CT
cleanout, which would have greatly delayed the entire operation. Having an available back-up well
14 SPE-184846-MS

for frac activities drastically reduced NPT thus greatly improving the overall operational efficiency
of the campaign. Figure 15 shows the schematic for the dual well frac rig-up.
3. Continuous water supply: Raageshwari field is located in a remote, arid region with continuous
water supply for frac activities being a key challenge. In previous campaigns, small water tankers (35
bbls capacity) were used to ferry water to the well pads. This method led to numerous water supply
interruptions resulting in serious NPT for the frac fleet. More than 8000 water loads would have been
required had this initial method been used in the 15 well campaigns. , hence pipelines were used to
carry water from a borewell in the production terminal to the various well pads. This eliminated all the
downtime associated with water supply. This solution also removed road traffic related HSE concerns
by reducing the number of truck loads hauled to the frac site every day.
4. Fracturing Fluid: The use of borehole water introduced a new problem. The borehole water was
saline, hard, and its boron content was above the permissible level for cross-linking of the base gel.
To use this water, a specially tailored frac fluid had to be developed. More than 60 lab tests were
conducted before obtaining an acceptable fluid formulation. Unfortunately, the new formulation had
much faster crosslink times which resulted in higher tubing friction and lower overall injection rates.
An in-well test with bottom hole gauges was eventually used to adequately characterize fluid friction
for design and evaluation purposes. These tests showed that the observed friction pressure was 20%
higher than expected (see Figure 16).
5. Perforating: India's regulations, restrict perforating to daylight hours, combined with the +3000m
depth of the wells, resulted in additional time and operation costs. In previous campaigns, the
perforation was performed using conventional E-line guns. This was acceptable when stages consisted
of single perforation intervals. However, it became untenable when limited entry and multistage job
designs required between 3 to 6 clusters per stage. Thus, the current designs required as many as 6
runs of conventional perforating guns. The combination of these factors resulted in executing more
than 1 frac in 2 days being virtually impossible. Selective firing switches which allowed up to 4 gun
firings selectively in a single run were introduced as a solution to this problem. This had a massive
impact on the frac schedule as perforating of all the clusters and pumping of the main frac could occur
on the same day.
6. Bridge plug milling: The various frac stages were isolated from each other using composite 10kpsi
bridge plugs. These bridge plugs were milled using CT with an expected milling time of less than 1
hour. In previous campaigns, the actual milling time exceeded 6 hours per plug. Considering there
were more than 5 bridge plugs in each well, this led to a considerable loss in efficiency as well as
contributing to the cost over-runs. After thoroughly reviewing the entire milling procedure and the
previous CT job charts, it was found that the CT and wellhead pressures were not kept constant as
per recommended procedures. This was due to the gas and condensate influx in the well from various
fracd zones. To avoid such fluctuations in the pressure, the milling procedure was modified and
returns were taken with constant back pressure through a surface well test spread. This reduced the
milling time to less than 1 hour per plug. This was a substantial improvement and contributed to the
improved overall efficiency of the campaign. Figure 17 shows job charts of milling operations before
optimization (first chart) and after optimization (second chart). The optimized milling time was less
than 1 hour per bridge plug.
7. Near zero gas and condensate flaring: Cairn India's goal is to eliminate gas and condensate flaring
when possible. As the location of the field is remote, previous well campaigns required some flaring
for well cleanup. This flaring was deemed unacceptable,and all the gas and condensate from the
respective gas and oil outlets in the separator were routed to the main production header via NRV set-
up. The separator pressure was used to push the gas and condensate into the header. It is estimated
that approximately 3 MMSCFD gas and 300 bpd condensate per day was saved using this approach
SPE-184846-MS 15

which contributed to the overall project economics. Figure 18 shows actual site pictures illustrating
the rig-up required to eliminate flaring during the milling process.

Figure 15Dual well frac rig up

Figure 16Comparison of estimated and actual tubing friction

16 SPE-184846-MS

Figure 17Comparison of pre and post optimization milling

Figure 18Site pictures illustrating the rig-up for the "flare free" milling operation

Fracture Diagnostics
While limited entry technique was considered as the option to increase the net pay interval covered by each
frac, it was also important to ensure its effectiveness. Various diagnostic tests were designed to cater to the
specific requirements for this field. These included:

Diagnostic Fracture Injection Test (DFIT)

Step Rate Test (SRT)/Step Down Tests (SDT)

Multiple Step Down test

SPE-184846-MS 17

Diagnostic Fracture Injection Test (DFIT)

DFITs were used to calibrate the minimum horizontal stress in the pay zones. After perforating the first
cluster of a zone, a small injection test above fracturing pressure was performed. The pressure decline was
monitored for an extended period. After the DFIT, the remaining clusters were perforated. The DFIT was
performed on single clusters to eliminate interference between clusters and the corresponding uncertainty
in analyzing the pressure decline.
Figure 19 shows the DFIT pressure decline analysis done for well P stage 3, utilizing G-function plot
(Denney, 2008). This methodology was applied for this field to estimate the closure stress and for qualitative
evaluation for pressure dependent leakoff (PDL). Table 3 shows details of the pumping operation and
the results from the analysis. In this case, the observed closure pressure was different from the lithostatic
minimum horizontal stress in the tested interval by around 200 psi. This information from all the DFITs was
utilized to calibrate the lithostatic stress in the pay intervals. The calibrated stress logs were then used for
the limited entry frac design. Note the indication of PDL between a G-time of 0 and 1.8. This signature was
not observed in every test but was generally indicative of higher leakoff and more productive formations.

Figure 19Well P, Stage 3, DFIT Analysis

Table 3DFIT Results Well P Stage-03

Description Unit Value Description Unit Value

Pumped Volume (bbls) 41 BH ISIP (psi) 7734

Pumping Time (min) 3.05 BH Closure (psi) 6895

Efficiency (%) 62 Closure G-Time 2.97

Table 4 shows a summary of all the DFIT analysis in the volcanics. It shows that the PDL signature
(indicating natural fractures) was seen more often in the Basalt and least often in the Felsic. Overall, only
39% of the cases had a PDL signature.

Table 4DFIT Analysis Summary

Volcanics Basalt Felsic

Total 54 100% 30 100% 24 100%

PDL 21 39% 13 43% 8 33%

HR 19 35% 9 30% 10 42%

Normal 14 26% 8 27% 6 25%

18 SPE-184846-MS

Step Down Tests

Once all the clusters were perforated, an SRT followed by an SDT step was conducted to evaluate the
perforation pressure drop and efficiency. In many cases this was combined with the cooling pad for the main
frac treatment to minimize cost and fluid in the formation.
As stated in Reservoir Stimulation 3rd edition, the SDT identifies and quantifies near wellbore tortuosity,
quantifies perforation effectiveness and can provide rough estimate of the number of perforations accepting
the fluid (Wright, 2000). The primary objective of these tests in this particular case was to determine whether
the actual perforation friction was sufficient to ensure diversion into all of the clusters.
Figure 20 shows an example SDT analysis from well P. It shows that about 2200 psi of total entry friction
and approximately 1430 psi of perforation friction pressure was observed at the maximum pumping rate
of 30 bpm. This is more than double the original estimate of perforation friction (600 psi), which indicates
reduced perforation efficiency. Thus, fewer perforations are taking fluid and there is room to increase the
number of perforations without adversely affecting diversion. Since this method of evaluation contains a
number of uncertainties and inherent assumptions, quantification and validation of this observation was
necessary before making final conclusions.

Figure 20Example SDT analysis from well P stage 3

While the tubing friction data provided by service companies is useful in predesign, it is often not
accurate enough for analysis purposes. This is especially true for treatments pumped down 3 inch tubing
where the tubing friction can be as high as 3000 psi and a 20% error in tubing friction would completely
mask the perforation friction. To remove this uncertainty, a series of SRTs with bottom hole gauges using
brine, 35 lb linear gel, and 35 lb crosslink were conducted. Figure 21 shows the results of these test and
were subsequently used to reanalyze all the SDTs. This eliminated the main uncertainty in the analysis of
perforation friction. This data was also invaluable in post frac history matching which will be discussed
later. Running a series of tests like this should be a high priority in any fracturing campaign and performed
as early in the campaign as possible.
SPE-184846-MS 19

Figure 21Results of tubing friction testing

At this point, it should be noted that SDT analysis theory assumes that injection is occurring into one
perforation set and one uniform zone. In the case of limited entry, multiple perforation clusters in multiple
layers with different stresses violates this assumption. In fact, since diversion into the zones is based on rate/
pressure drop, it is likely that during the step-down procedure, the % rate distribution will change as the
rate is decreased and at some point, one or more clusters in the higher stress layers could stop taking fluid
or even start cross flowing into the other clusters. This phenomenon would probably appear as increased
tortuosity in the analysis. As a result, by the end of the campaign, we decided to assume that the total entry
friction was the perforation friction. For diversion purposes, it is the total perforation pressure drop at any
rate that matters.

Multiple Step Down Tests

Multiple sequential SDTs was another method used to determine the perforation efficiency. This technique
was employed in the 2nd stage of well J. First, three of the four clusters were perforated with 8 holes each
and an SRT was run to determine the number of open holes. This was followed by perforation of the fourth
cluster and a second SRT. Figure 22 shows the plot of total entry friction vs rate for each injection. It shows
that the entry friction at 30 bpm dropped about 900 psi. Since the both SRTs were run at similar rates
with linear gel, the only difference would be the additional perforation area. Analysis of this combined test
indicated that 50 to 60% of the perforations were taking fluid. A temperature survey (see Figure 23) after
the final SRT confirmed that all four clusters were taking fluid.

Figure 22Well J, sequential SRT

20 SPE-184846-MS

Figure 23Well J Post 2nd SRT Temperature Survey

Evaluation such as these in more than 35 tests consistently showed higher entry friction. This provided
the basis for increasing the maximum number of perforations from 24 to 30. Eventually as many as 6 clusters
per stage were successfully placed. Entry into all the clusters was confirmed with temperature surveys
performed after SRTs and frac jobs like the post SRT survey done in well T (see Figure 24).

Figure 24Well T, SRT Simulation vs post SRT temperature survey with 6 clusters

Post Treatment Pumping History Matching

The following is an abbreviated version of the procedure used to history match each treatment:
1. Transfer actual job data to the fracture simulator.
2. Run a baseline simulation using the log data and the measured friction pressure.
3. If the measured instantaneous shut in pressure (ISIP) is lower than the simulated ISIP, then shift the
stresses to match the measured ISIP.
4. If the measured ISIP is higher than the simulated ISIP and if the post frac decline rate is steeper than
the simulated decline rate, adjust the fluid loss coefficient.
5. Match the pad surface pressure by changing the number of open perforation.
6. Match the proppant stage surface pressures by adjusting the proppant friction coefficient.
SPE-184846-MS 21

The model was run in either Pseudo 3D (P3D) or fully 3D mode depending on the case. Normally
perforation erosion was turned on. While the perforation erosion calculated by the model appears to happen
much more quickly than what we see in actual jobs, the total reduction in pressure due to erosion is consistent
with the observed data.

Well V: 4 Cluster Example

Well V stage 2 is a good example of where 4 clusters were used to access 4 different pay sections. This was
one of the deeper zones near the bottom of the Basalt. The treatment consisted of an SRT followed by the
main frac on the next day. The ISIP from the SRT was 4100 psi. The main treatment consisted of about 2150
bbls of crosslink and 241 Klbs of proppant (7 Klbs 100 mesh ISP, 117Klbs 20/40 ISP and 117 Klbs 16/30
ISP). The maximum proppant concentration was 8 PPA and the treatment was pumped as per plan. The ISIP
of the main treatment was 5100 psi. Thus, a 1000 psi increase in net pressure was observed between the
SRT and the main frac job. The surface pressure history match of the actual pumping schedule is shown in
Figure 25. This match was achieved using the measured friction pressures, a slurry friction factor of 1.75
(Hannah, 1983), perforation efficiency of 67%, and an average fluid loss coefficient of 0.0001 ft/sqrt(min).
No adjustments to the stress profile were required to match the data. It reflects a good correlation with the
available data and a good match of the ISIP. The pressure drop which occurs in the simulation at 30 minutes
is due to a reduction in perforation friction caused by perforation erosion which is modeled by the fracture
simulator (Crump, 1988). Figure 26 shows the diagnostic plot (Nolte & Smith, 1981) for the simulation. It
indicates a moderately contained frac and is consistent with the 1000 psi gain in ISIP. Figure 27 shows a
merging of the post frac temperature survey, fracture width at closure and the predicted conductivity profile.
The observed height from the temperature survey is consistent with the simulated height.

Figure 25Well V history match surface pressure

22 SPE-184846-MS

Figure 26Well V Nolte-Smith plot

Figure 27Well V fracture simulation results with temperature survey

Well W: 5 Cluster Example

Well W stage 2 is a good example where 5 clusters were used to access 5 different pay section. This stage
was completed using channel fracturing technology (Medvedev, Kraemer, Pena, & Panga, 2013). This was
one of the deeper zones near the bottom of the Basalt. The first attempt to fracture this stage was aborted
near the end of the job due to a blender issue after 80 Klbs of proppant was pumped. The ISIP from this first
attempt was 4261 psi. As this treatment incorporated channel fracturing and the tail-in was not pumped,
it was decided to refrac the stage (Barasia, 2014). The refrac of this stage began with an SRT the night
before the main frac to ensure that injectivity could be established. The ISIP from the SRT was 4251 psi
which was quite close to the ISIP from the original frac. The refrac treatment consisted of about 2230 bbls
of crosslink and 158 Klbs of proppant (75Klbs 20/40 ISP and 83 Klbs 16/30 ISP). The maximum proppant
concentration was 8 PPA and the treatment was pumped as per plan. The ISIP of the main treatment was 4482
psi. The surface pressure history match of the actual pumping schedule is shown in Figure 28. This match
was achieved using the measured friction pressures, a slurry friction factor of 0.3, perforation efficiency
of 58%, and an average fluid loss coefficient of 0.0003 ft/sqrt(min). A -300 psi adjustment to the stress
profile was required to match the data. This match is not as good as the other matches. This difference
may be due to the previous frac treatment or friction differences due to the channel fracturing. Figure 29
shows the Nolte-Smith plot for the simulation. Figure 30 shows the width profile, post frac temperature
survey and conductivity profile from the simulation. The temperature survey shows that all the clusters took
SPE-184846-MS 23

a significant amount of fluid and the upper 2 clusters converged into one fracture. The top of the fracture
from the temperature survey appears to be 10 to 20 m above the simulated top of fracture.

Figure 28Well W Surface pressure history match

Figure 29Well W Nolte-Smith plot

Figure 30Well W width profile, post frac temp survey & conductivity profile

Well T: 6 Cluster Example

Well T stage 7 is a good example of 6 clusters used to access 6 different pay section. This was the uppermost
stage and completed in the top of the Basalt. The treatment consisted of an SRT followed by a temperature
survey the night before the main frac. Fluid entry into all perforation was confirmed by the post SRT
temperature shown in Figure 24. The ISIP from the SRT was 4041 psi. The main treatment consisted of
24 SPE-184846-MS

about 2500 bbls of crosslink and 363 Klbs of proppant (137Klbs 20/40 ISP and 226 Klbs 16/30 ISP). The
maximum proppant concentration was 7.7 PPA and the treatment was pumped as per plan. The ISIP of the
main treatment was 5500 psi. Thus, a 1400 psi increase in net pressure was observed between the SRT and
the main frac job. The surface pressure history match of the actual pumping schedule is shown in Figure 31.
This match was achieved using the measured friction pressures, a slurry friction factor of 1.6, perforation
efficiency of 47%, and an average fluid loss coefficient of 0.0019 ft/sqrt(min). No adjustments to the stress
profile were required to match the data. It shows a good agreement with the available data and a good
match of the ISIP. The pressure drop which occurs at 40 minutes is due to a reduction in perforation friction
caused by perforation erosion which is modeled by the fracture simulator. Figure 32 shows the Nolte-Smith
diagnostic plot for the simulation. It shows that a slight tip screen out (TSO) (Smith M.B., 2000) was
achieved at the end of the pumping and is consistent with the 1400 psi gain in ISIP. Figure 33 shows the
conductivity profile from the simulation. It shows that all the clusters took a significant amount of fluid and
the upper 4 clusters converged into one fracture. The lower two clusters converged into a second fracture.

Figure 31Well T surface pressure history match

Figure 32Well T Nolte-Smith plot

SPE-184846-MS 25

Figure 33Well T conductivity profile

Summary of History Matching

Two key history matching points are the fracture height and the ISIP. Since most of the wells had significant
proppant fill after the frac, it was impossible to determine the bottom of the lowest fracture from the
temperature logs. However, in many cases it was possible to locate the top of the upper most frac. Figure
34 shows a crossplot of measured vs actual fracture top (normalized to a zero depth). The closer the points
are to the diagonal line, the better the match. The prediction of the top of the fracture was very successful.

Figure 34Cross plot of frac top from temperature survey and frac simulator

Since the ISIP could be measured in almost all of the treatments, more match points were available.
Figure 35 shows a cross plot of the measured ISIP vs the simulated ISIP. Once again, the closer the points are
to the diagonal, the better the match. There is very good agreement between the actual and simulated ISIPs.
26 SPE-184846-MS

Figure 35Cross plot of measured ISIP vs history matched ISIP

Table 5 shows a summary of the history matched parameters. Some of the key points are:
1. Stress had to be shifted in about 70% of the cases. The average shift was about 300 psi which is less
than a 4% change.
2. The average error in matched ISIP was only 3%.
3. The average slurry friction factor was 1.30.
4. The average overall friction factor (changes to base friction) was 1 (meaning almost no changes were
made to the fluid friction in the simulation.

Table 5Summary of history matching

Parameter Unit Min Max Average

Total Frac Height M 13 134 70.4

Active Perforations Number 7 60 18.4

Number of Clusters Number 1.0 6.0 3.4

% Error in ISIP Match % 0% 18% 3%

Abs Value Stress Shift Psi 0 3600 315

Leakoff Coeff ft/sqrt(min) 0.0003 0.0035 0.0016

Slurry Friction Factor 0 2 1.3

Overall Friction Factor 0.9 1.2 1.0

Flowback Results
During the appraisal phase, the objective of the flowback was to establish the initial productivity (IP) of the
wells. During the development phase, in the most recent campaign, the objective of flowback was limited
to initial clean up and getting the total water production low enough to send to the production header.
Additional tests were performed to evaluate the effect of commingling multiple zones on the productivity of
each zone. After fracturing, the bridge plugs were milled and , the wells were flown through a separator until
the WGR reduced to about 30 bbl/MMSCFD which was the cutoff for the production header. In general,
this WGR was achieved within 8 to 10 days during which time approximately 20% (see Figure 38) of the
frac fluid was recovered.
The flow back graphs in Figure 36 show the difference between the testing of the Basalt (middle
formation) vs the commingled testing with all three formations Basalt, Felsic and Fatehgarh. While the
SPE-184846-MS 27

addition of the other zones increased deliverability, it is impossible to know the distribution of the added
flow.

Figure 36Flowback of Well J for individual zones and commingled zones

Time lapsed production logging (a series of production logs run over a few months) was used to
understand the impact of commingled production on individual formation performance for a number of
wells. Figure 37 shows 5 sequential production logs for well P. The first log was taken during the first
flow period with only the Felsic open. The well was flowed on a 16/64th choke with a surface pressure of
2970 psi with an average bhfp of 4225 psi. It shows good contribution from each of the Felsic frac stages
with about 70% of the production coming from the top two stages. The second production log (also only
in the Felsic) was taken 3 months later and at a higher rate. The well was tested on a 17.5/64 choke with
at a surface pressure of 2900 psi and a bottom hole producing pressure of 4150 psi. It reflects an improved
profile with only about 40% of the production coming from the top two stages and 60% from the bottom
two. Note, however, it is unclear whether the increased rate or time has improved flow distribution. In the
third production log, , all the zones (Felsic, Basalt, and Fatehgarh) have been opened. While all the stages are
contributing, the lowest two stages and the top stage are under performing. By the 4th production, the lowest
stages have picked up and this is at a lower overall production rate. The 5th and final production log was run
at a higher rate (drawdown) two months after the 4th production log, and more balanced flow is seen from all
the stages. This series of production logs clearly demonstrate that a) all stages are producing, b) contribution
improves with production rate (increased drawdown) c) contribution improves with production time (or
cleanup) and d) 100% of the stages and approximately 90% of the clusters are contributing. Production logs
run in 16 out of the 20 wells with limited entry (93 stages and 297 clusters) show that 100% of the stages
are contributing and about 80% of the clusters are producing to some degree.
28 SPE-184846-MS

Figure 37Time Lapsed Production Logs for Well P

Figure 38Percentage Frac fluid recovery with vs flow back days

While evaluating the early gas production data to evaluate the well performance, fracturing fluid recovery
varied from 9% to 29% during initial flow period which lasted from 9 to 23 days. The fluid flowback water
recovery was not impacted by shutin time, mass of proppant pumped, total volume of frac fluid or rate at
which job was pumped. The data for all wells with some flowback information is summarized in Table 6.
SPE-184846-MS 29

Table 6Summary of Flowback Results for 14 Wells

Well Total Frac Fluid Cum. Gas vol Cum Water % Frac Final Total
(bblsx100) (mm SCF) (bbl 100) Fluid WGR Test Days
recovery

N 178.0 55.1 15.2 9%

O 157.0 59.4 14.0 9% 14.0 9.7

S 96.0 27.9 15.0 16%

V 168.4 26.1 26.9 16% 84.0 6.5

L 184.0 104.6 31.0 17% 20.0 15.5

AA 123.0 54.6 21.8 18% 19.0 17.4

Y 163.0 89.2 29.4 18%

Z 125.0 33.1 23.9 19% 44.0 9.0

K 135.0 135.9 27.9 21% 10.0 20.4

X 145.0 128.0 36.4 25% 15.0 22.5

U 170.0 29.7 44.1 26% 42.0 11.5

AC 107.7 19.6 28.8 27% 77.4 7.2

M 120.0 15.7 35.0 29% 155.0 8.4

W 103.0 63.3 81.7 79% 53.0 23.7

Figure 38 shows a graph of the final frac fluid recovery vs the number of days the well was flown through
the separator. Note that each point is a separate well. Figure 39 shows a graph of WGR vs flowback time.
Together the two plots indicate that there are three distinct types of wells.

Figure 39WGR vs flowback days for 11 wells

The first group (represented by the triangles) are wells with high initial water production returning 20%
of the frac fluid very quickly after which the water quickly drops off and cumulative recovery remains low.
The wells which have shown PDL during DFIT analysis were part of first group of wells. These wells also
had the highest initial productivity. Thus lower and quicker frac fluid recovery trends and DFIT decline
analysis for wells with PDL were coming from the same set of high IP wells. This observation led to the
30 SPE-184846-MS

conclusion that wells with presence of natural fractures are having the higher initial production rate. An
attempt was made to maximize the production from natural fractures by propping them with 100 mesh ISP.
The wells which utilized 100 mesh ISP are still under evaluation.
The second group (represented by blue circles) are moderate water recovery wells where the 20% or
more frac fluid recovery follows a linear trend but at a much lower rate than group 1. The third group
consisting of 2 wells in this dataset are outlayers (represented by red squares). They returned the most water
and maintained high WGR. It is likely that these wells were producing some formation water in addition
to the frac fluid.

Cost Optimization
Improvements in RGD economics were achieved by minimizing the cost of the operations while maximizing
the well deliverability and EUR. This section will review the 50% reduction in fracturing costs from the
last two campaigns. After the 5 well campaign, a thorough analysis identified some key areas where cost
reductions could be achieved. These areas included:
1. Contractual Savings
2. Operational Efficiency
3. Design Optimizations
An in-house cost tracking tool was developed and used to ensure that all operational and design related
decisions had a positive impact on project economics. The items which had the most impact on the project
economics are listed below:
1. Contractual Savings
a. Reduced supply cost (chemicals and proppant) achieved through a better understanding of the
project requirements and ensuring that all key products were specified in the scope of work.
b. Reduced service cost.
2. Operational Efficiency
a. Dedicated water supply to well pads.
b. Having adequate site storage of proppant and mobile silos to ensure an uninterrupted proppant
supply at the well-site.
c. Rigging up of multiple wells for frac activities.
d. Ensuring high level of simultaneous operations on site post appropriate risk assessment.
e. Ensuring all services on site have a dedicated maintenance slot to avoid any equipment
breakdowns.
f. Reduced perforating runs by use of selective firing switches for perforation.
g. Reduced milling time by improved coil tubing milling procedure.
3. Design Optimization
a. Usage of cooling pad linear gel for step rate tests.
b. Temperature surveys were run at night and only when they did not interfere with/delay other
operations.
c. Increasing number of clusters per stage and therefore reducing total number of stages per well.
d. Reduction in screen-out rate to less than 5% through improved model calibration.
Figure 40 shows how the cost per frac evolved from the initial 5 well campaign to the 15 well campaign.
Furthermore, based on optimizations in the 15 well campaign, an additional 19% reduction in cost/frac is
SPE-184846-MS 31

expected in the upcoming 37 well project. Figure 41 shows a segmentation of the total costs incurred in the
15 well program. This type of segmentation helps identify key areas for focus and future cost reduction.

Figure 40Evolution of cost savings in RDG

Figure 41Cost segmentation in the RDG 15 well frac campaign

Conclusions
The following conclusions were drawn from the fracturing campaigns:
1. Successful hydraulic fracturing of the RDG volcanics has been proven with more than 100 stages.
2. Limited entry offers an economical way to maximize reservoir contact in RDG volcanics.
a. Temperature logs show all clusters are being fractured.
b. Production logs show that at least 80% of the clusters, (but all stages), are producing and
contribution increases with time.
c. The successful application of Limited Entry is dependent on mainting the pressure drop
throughout the job. For this reason, perforation erosion must be taken into consideration.
d. When properly designed, as many as 6 perforation clusters can be treated at one time.
32 SPE-184846-MS

3. Conventional fracturing models are adequate for designing and evaluating treatments in the RDG
volcanics.
4. A focus on operational efficiency can have a big impact on overall costs. In the case of RDG the cost/
frac was reduced by 50% between the 5 well and the 15 well campaigns.
5. The optimized use of breakers is essential for maximum fluid flow-back after the treatment. It can
significantly impact production rates and recovery.
6. Net pressure match and time lapsed post frac production logs are recommended to optimize the future
job design (size) and to further improve the reservoir performance.
7. An in house work flow was developed to identify sweet spots for perforating and predicting the ISIP
within 3-4 % of variation.
8. The wells with presence of natural fractures have shown high initial production and low recovery of
frac fluid.

References
Arora, G., Stolyorov, S., Gupta, A., Purusharty, N., R., S., & Mathur, M. (2011). The Application Of Sand Jet Perforation
In Multistage Hydraulic Fracturing Treatments In Volcanic Reservoir India. SPE 143423-MS. Society of Petroleum
Engineers.
Barasia, A. S. (2014). Tail-in proppant and it's importance in channel fracturing. SPE 169227-MS. Society of Petroleum
Engineers.
Bineet Mund, R. S. (2016, December 16). An Integrated Approach for Facies Characterization in Raageshwari Deep Gas
Volcanic Reservoirs, Barmer Basin, India. Petrotech. Delhi, NCR, India: Petrotech.
Crump, J. a. (1988, August 1). SPE 15474: Effects of Perforation- Entry Friction on Bottomhole Treating Analysis. Journal
of Petroleum Technology, 10411048.
Davis, E. R. (1997, June 1). Interpretation of Fracture Height From Temperature Logs - The Effect of Wellbore/Fracture
Separation. . 29588-PA. Society of Petroleum Engineers.
Denney, D. (2008, March 1). Hydraulic Fracturing: Holistic Fracture Diagnostics. Society of Petroleum Engineers.
JPT(March).
Dobkins, T. (1981). Improved Methods to Determine Hydraulic Fracture Height. Journal of Petroleum Technology, 719.
Hannah, R. R. (1983, January 1). Time Calculation of Accurate Bottomhole Fracturing Pressure From Surface
Measurements Using Measured Pressures as a Base. 12062-MS. Society of Petroleum Engineers.
Jha, M. S. (2010, January 1). Monobore Design Optimises Slimhole Raageshwari Deep Gas Development. 128394-MS.
Society of Petroleum Engineers.
Lagrone, K., & Rasmussen, J. (1963, July 1). A New Development in Completion Methods- The Limited Entry Technique.
SPE 530-PA. Society of Petroleum Engineers.
Medvedev, A. V., Kraemer, C. C., Pena, A. A., & Panga, M. K. (2013). On The Mechanisms of Channel Fracturing.
163836-MS. Society of Petroleum Engineers.
Nolte, K., & Smith, M. (1981, September). Interpretation of Fracturing Pressures. Journal of Petroleum Technology,
1767-1775.
Schumberger Oilfield Services i-handbook. (version 1,0.4,5). Field Data Handbook - i Handbook.
Shaoul, J. R. (2007, January 1). Massive Hydraulic Fracturing Unlocks Deep Tight Gas Reserves in India. SPE 107337-
MS. Society of Petroleum Engineers.
Smith M.B., J. W. (2000). Basics of hydaulic fracturing. In M. E. Nolte, Reservoir Stimulation (3rd ed., pp. 5-3). Wiley.
Vermani, S. G. (2010). Fracturing Deep Nonconventional Volcanic Reservoir: A Case History Raageshwari Gas Field,
Onshore India. SPE 132932-MS. Society of Petroleum Engineers.
Wright, C. (2000). Rate step-down test analysisa diagnostic. In M. E. Nolte, Reservoir Stimulation (3rd Edition) (3rd
ed., pp. 9-32). Wiley.