SPE 163117

High Rate ESP Application in Onshore Rajasthan Fields: A Case Study

M. Jha, SPE, R. Singh, SPE, C. Chavan, SPE, and R. Karthik, SPE (Cairn India Limited), A. Simpson (Triangle
Engineering), J. Sudharshana, SPE (Baker Hughes)

Copyright 2012, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Artificial Lift Conference and Exhibition held in Manama, Bahrain, 27-28 November 2012.

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

This paper describes the selection, design, successful application and performance monitoring of Electrical Submersible
Pumps (ESP) in the giant Mangala oil field and Thumbli water field situated in the Barmer basin in Rajasthan, India.
Mangala oil field contains in excess of 1 billion barrels of STOIIP (Stock Tank Oil Initially in Place) in high-quality fluvial
reservoirs. The field was brought on production in August 2009 and is currently producing at the plateau production rate
of 150,000 bopd of which approximately more than one- third of the oil production is from the ESP oil wells.

To support the water requirement of Mangala and other satellite oil fields, Thumbli source water field was developed with
5 water production wells with up to 4 wells operating at a time. Each of these water wells is installed with 60,000 bwpd
capacity pumps and the field is currently producing up to 225,000 bwpd to meet the water requirements of Mangala and
other satellite fields.

The Mangala oil field is a multilayer, multi-Darcy reservoir, has waxy viscous crude with in-situ oil viscosity up to 22 cp
and wax content in the range of 18 to 26%. The field was developed using hot water flood for pressure maintenance.
Significant production challenges included unfavorable mobility ratio with early water cut and hence the early
requirement of artificial lift to maintain the plateau production rate. The field has 12 horizontal producers and 92 deviated
producers. ESP was selected as the artificial lift method for the high rate horizontal producers while hot water jet
pumping was selected as the artificial lift method for low rate deviated oil wells. Each horizontal well is capable of
producing up to 15,000 blpd and high rate ESPs were designed and installed to deliver the production requirement.
Currently 8 of the 12 horizontal producers are on ESP lift and the remaining four wells are planned for ESP installation in
the near future. Apart from two early ESP failures during installation, ESPs have had a good run life; the paper also
describes lessons learnt from the infant mortalities.

The Thumbli water field, located ~20 km southeast of Mangala field has been developed to meet the water requirement
of Mangala and other satellite fields. Thumbli water aquifer is a shallow water field which contains water of ~ 5000 ppm
salinity with dissolved CO2, oxygen, chlorides and sulphate reducing bacteria (SRB). 5 high capacity water wells were
drilled in Thumbli field to meet the huge water demand from Mangala for water injection in Mangala and satellite field
injector wells, hot water circulation in oil production wells and associated water requirement for boilers etc. 1000 HP
water well ESPs were designed to produce up to 60,000 bwpd from each well with installed water production capacity of
up to 300,000 bwpd from Thumbli field.

A state of the art ESP control and monitoring architecture including ESP tornado plotting was developed and
successfully implemented in the ICSS to remotely operate, monitor and optimize ESP well performance from the central
control room within Mangala field and from the company headquarter located in Gurgaon.
2 SPE 163117

Introduction

The Mangala field was discovered in January 2004 by the discovery well N-B-1 in the license block RJ-ON-90/1 (Figure-
1). Mangala was continually appraised during 2004-2007 with the drilling of eight more wells, the acquisition and
interpretation of a 3D seismic survey, and major data gathering and analysis efforts involving core, fluids, and well
testing.

The main reservoir unit in Mangala is the Fatehgarh Group, which has been sub-divided into the Lower Fatehgarh
Formation dominated by well-connected sheet flood and braided channel sands, and the Upper Fatehgarh Formation
dominated by sinuous, meandering, fluvial channel sands. Five reservoir units (FM1-FM5) have been named from the
top down. Fatehgarh sand properties are excellent, with porosities of 21-28% and permeability of 200-20,000 milliDarcies
(mD). Average permeability is ~5 Darcy. The Mangala structure is a simple tilted fault block dipping at ~9º to the
southeast. The depth of the Mangala structural crest at the Fatehgarh level is ~600m sub surface and the oil-water
contact (OWC) is at 960m sub surface, resulting in a total oil column of ~360m. Ground level elevations range from 175-
200m above MSL. (Figure-2)

Mangala contains waxy sweet crude oil with API gravity ranging from 20ºAPI near the OWC to 28ºAPI higher in the oil
column (average ~27ºAPI). The crude has an in-situ oil viscosity of 9-22 centipoise (cp), with live oil wax appearance
temperature (WAT) ~6oC lower than the average reservoir temperature of 65oC. The oil also has a high pour point of 40-
45oC, which had a great influence on all aspects of the project and process design including the ESP well design.

Due to a small mapped aquifer, primary recovery is estimated at less than 10 % STOIIP with simple depletion (small
mapped aquifer). The development plan involved implementing a hot water flood from production startup to maintain
reservoir pressure and effectively sweep reserves. Water heated to temperatures above the WAT is being injected to
ensure injectivity and efficient oil sweep. A suitable water source was identified for water flood operations following due
deliberations. Western Rajasthan is a drought-prone area with limited fresh water supplies. It was realized from the
outset of development planning that any water produced for injection into the oilfields should have a negligible impact on
water resources of value to the local community.

Fortunately, the fields’ water requirements could be met by a saline water source (TDS > 5,500 ppm) that is of no
possible use for either drinking or agricultural use. A series of studies concluded that the use of saline ground-water was
the preferred solution and the regional Thumbli Sand aquifer was identified as the optimum source of water due to its:

• Vast saline water reserves

• Excellent aquifer properties
• Shallow depth (0-600m Below Ground Level) of location
• Relative proximity to the oilfields, and
• Relative distance from freshwater well fields.

A schematic showing the main elements of Mangala field and Thumbli aquifer is shown in Figure 3.

Horizontal wells completed with stand alone screens (SAS) and inflow control devices (ICDs) are planned for the blocky
FM3 and FM4 units and these wells form an important part of maintaining the initial plateau production and extracting
reserves in an optimized fashion. 12 horizontal wells have been drilled and completed in the FM3 and FM4 units. Each of
these horizontal well was drilled with a target production rate of up to 15000 blpd. Due to the high volume lift
requirement, absence of lift gas and flow assurance issues related to gas lifting viscous oil, ESPs were selected as the
artificial lift for Mangala horizontal wells. (Figure-4)

Thumbli water well field was developed with five wells, with each well spaced 100 meters apart in a linear pattern. Each
well was targeted to produce up to 60,000 bwpd and ESP was the natural choice of artificial lift for this isolated water
field. (Figure-5)

Mangala Horizontal Well ESP Design

Given the nature of development and the possibility of upfront ESP requirement to sustain production from the
field, it was decided to design and specify ESPs based on the best available engineering estimates about the
horizontal wells.
SPE 163117 3

Productivity Index. No horizontal wells were drilled previously in the field. Based on the available understanding
of FM3 and FM4 sand, in which the horizontal wells were planned, and using industry standard methods of
estimating PI of the horizontal wells, it was predicted that the productivity index of the horizontal wells could range
between 70 to 125 stb/d/psi with base case PI of 100 stb/d/psi.

Production Rate. To support the plateau production from Mangala field, the horizontal wells were identified as key
wells with target liquid production of 10,000 blpd, design sensitivities were run for low case of 7500 blpd and
upside of 20000 blpd.

Emulsion. Mangala crude was expected to be prone to emulsion formation based on all the laboratory work done
at the time of designing ESPs. Emulsion modeling was considered critical for ESP design and it was predicted that
the emulsion viscosity will be 50% of the values observed during the laboratory work under similar shear conditions
as (Figure-6):

a) The laboratory emulsion work was done using dead oil and it was anticipated that the degree of emulsion
formation will be lower in case of live crude.
b) The well completion includes a dedicated chemical injection line for dosing demulsifier upstream of the
ESP which could be used in case emulsion formation starts.

The predicted reduction in emulsion viscosity under field conditions was also based on analogue field behavior.

Water Cut. Based on reservoir simulation data and vertical lift modeling it was estimated that the horizontal wells
will require artificial lift to maintain production after approximately 40 % water cut.

Considering all the sensitivities and possible production scenarios, it was decided to size ESP as four broad types,
A, B, C and D with the design limits as specified in Table 01.

ESP Designation Nominal HP Nominal Duty Min/ Max Flow Additional Conditions
Operating Point (blpd) Range(blpd)
Type ‘A’ 320 10,000 7,500/ 15,000 50% water cut
Type ‘B’ 460 10,000 7,500/ 15,000 70% water cut
Type ‘C’ 460 15,000 12.000/20,000 25% water cut
Type ‘D’ 600 15,000 12,000/20,000 50% water cut

Table 1: ESP sizing scenarios for Mangala Horizontal Wells

Based on the design work and on further analysis of the well production behavior, it was decided to specify Type B
ESPs in all wells as the base case and conduct design changes in future, if warranted, in case the horizontal wells
behave in a manner different than initially expected.

In addition to the flow range and HP requirement, a detailed analysis of the particle size distribution and geometry
of sand from various core samples was conducted and it was decided to run radial support bearing at every
second stage of the pump. In addition to these, Mangala crude had high CO2 concentration with up to 25 mol % of
CO2 measured in the gas phase. It was decided that all hydrocarbon flow wet downhole equipment including ESP
components will have 13Cr80 or equivalent metallurgy.

The downhole equipment specifications of Type ‘B’ pump is shown in Table 02 below:
4 SPE 163117

Item Description
Pump 675 series, 22 stage HC20000, 1:2 AR housed pump with 13Cr housing and 416 stainless steel
Head and base, bolt on discharge, with integral intake and 513 base.
UT Seal 513 series, upper tandem GSB3 DB seal, bag/bag/labyrinth construction with 13Cr housing and
416 stainless steel head and base, Aflas bag, Aflas premium face seal, high load thrust bearing,
high strength shaft material.
LT Seal 513 series, lower tandem GSB3 DB seal, bag/bag/labyrinth construction with 13Cr housing and
416 stainless steel head and base, Aflas bag, Aflas premium face seal, high load thrust bearing,
high strength shaft material.
Motor 562 series, single 460HP 2477V/153A 60Hz rating KMH-J motor, 13Cr housing with 416 stainless
steel head and base, and high temperature construction and oil.
Cable No.2 CENR 5 KV, Galvanized Steel Armor

Table 02: Type ‘B’ Downhole Equipment Specification

A detailed analysis was completed to show the effect of motor operating temperature on seal expansion capacity.
The seal specified, tandem 513 series GSB3 DB was modeled in Autograph PC® and it was found that the upper
chamber, closest to the pump, exceeds 100% expansion capacity at a motor operating temperature of 385°F
against normal motor operating temperature of 200°F. The simulation feature of Autograph PC® was then utilized
to estimate time required for the motor operating temperature to reach 385 °F upon initiation of a deadhead
condition. With the case of an ESP packer in the well, it was estimated that it will take ± 16 minutes for the motor
temperature to increase to 385 °F. Obviously, these were considered estimates and served to highlight the danger
of a deadhead condition as well as the fact that one cannot rely on a “time factor” as an alarm. Proper alarms
based not only on motor operating temperature but also pump intake and discharge pressures and various other
surface parameters were considered; these are discussed separately in the ESP control section.

Additional analyses were conducted on the seal thrust bearing loading limits, motor temperature limits, shaft
loading limit, motor voltage and current loading limit and low flow past motor limit to clearly identify the operating
envelope for the ESP system.

Accidental dead head scenarios were also simulated to understand the behavior of complete well system under
the following conditions:

1. Wellhead maximum pressure in the event that the surface valve is closed at the defined operating point.
2. Wellhead maximum pressure in the event that the system is started against a closed surface valve.
3. Maximum differential pressure in the event that the system is started against a closed surface valve.

The surface flow line and X-mas tree pressure rating was up to 3000 psi due to the requirement of jet pumping in
the deviated wells and the ESP housing differential pressure rating was 2680 psi. It was established by calculation,
that these limits are not exceeded under any foreseeable operating scenarios.

Thumbli Water Well ESP Design

Thumbli ESP design is based on the experience of three previous water wells drilled in the Thumbli area, proposed
improvements in drilling design to minimize formation damage along with an additional plan for insitu natural gravel
pack development. It was estimated that the new Thumbli water wells will have PI of ~750 bpd/psi with static liquid
level of ~ 40 m BGL (Below Ground Level). The expected delivery rate of 60,000 bwpd was required of each well
at surface pressure of 350 psi. One of the important design aspects was that the Thumbli- MPT pipeline has a
design pressure rating of 750 psi and it was decided to keep the deadhead pressure below this value by careful
selection of pump head capacity curve. The ESP equipment selected for Thumbli water wells is shown in Table 03.

Item Description
Pump 900 series, 15 stage WNE1600, 1:2 AR bolted bowl construction with nickel aluminum bronze
(NAB) stages.
UT Seal 675 series, upper tandem HSB3 DB seal, bag/bag/labyrinth construction with 9Cr 1 Mo housing
coated with 6 CRC and 416 stainless steel head and base, Aflas bag, Aflas premium face seal,
high load thrust bearing, high strength shaft material.
LT Seal 675 series, upper tandem HSB3 DB seal, bag/bag/labyrinth construction with 9Cr 1 Mo housing
SPE 163117 5

coated with 6 CRC coating and 416 stainless steel head and base, Aflas bag, Aflas premium face
seal, high load thrust bearing, high strength shaft material.
UT Motor 725 series, upper tandem 500HP 1965V/153A 60 Hz rating HMI motor, 13Cr housing with 416
stainless steel head and base.
LT Motor 725 series, lower tandem 500HP 1965V/153A 60 Hz rating HMI motor, 13Cr housing with 416
stainless steel head and base.
Cable No.1 CENR 5 KV, Stainless Steel Armor

Table 03: Thumbli Water Well Downhole Equipment Specification

Metallurgy Selection. Because of the high amount of dissolved oxygen in Thumbli water and moderate salinity,
the pump metallurgy was chosen as Nickel Aluminum Bronze (NAB) which is known to have excellent performance
in high dissolved oxygen environment and hence finds extensive usage in marine environment such as ship
propellers. The initial idea was to select 316SS or 25 Cr SS but NAB was chosen as this was a standard material
with proven track record whereas 316 SS or 25 Cr SS were considered non- standard ESP metallurgy which could
have added to the delivery schedule. The motor and the seal housing were constructed of 9Cr1 Mo with 6 mills of
additional CRC (Corrosion Resistant Coating).

Equipment Limit Specification. Similar to Mangala horizontal wells, a detailed analysis was conducted to
understand the selected equipment operating envelope. This involved understanding the seal expansion capacity,
thrust bearing loading limits, motor temperature limits, operating current and voltage limits, shaft loading limits,
dead head pressure and ESP housing pressure limits.

Mangala EOR Well ESP design

Very early in the life of Mangala field, it was decided to run an Enhanced Oil Recovery (EOR) pilot with inverted
five spot patterns. The EOR pilot producer well was required to have a bypass leg to facilitate frequent logging
requirement across the sand face for better understanding of the pilot. One of the key problems anticipated with
the EOR producer was low fluid velocity past the motor because of low production rate of EOR producer with
nominal 1500 blpd through a relatively large production casing size of 9-5/8” (Figure: 7). This translated to a fluid
velocity of 0.384 ft/sec past the motor which was far lower than industry accepted rule-of-thumb of 1 ft/sec for
effective motor cooling. Various options including the provision of shunt tubes, motor shroud and the possibility to
increase the motor size etc. were considered in addition to selecting a higher temperature rating motor. Finally
based on careful ESP performance analysis using Autograph PC®, it was decided to run a higher temperature
rating motor with expected motor operating temperature of 203o F. During operations the motor temperature has
remained below 195o F thus validating the design work. The downhole equipment specification of the EOR ESP is
shown in Table 04 below:

Item Series Description

25 Stage, 400 P35 SSD with 13 Cr housing metallurgy, and AR (abrasion resistant) trim
Pump 400 pump.
Tandem seal with high load thrust bearing, Inconel shaft, 13 Cr housing metallurgy, Aflas
Seal 400 elastomer bags, and B/B/L/B/B/L construction.
Motor 450 42 HP motor with 13 Cr housing. Temperature Rating 325 F
Cable No.2 No.2 CENR 5 KV, Galvanized Steel Armor
By pass
System NA 3-1/2” VAM FJL Bypass system complete with telescopic swivel nipple

Table 04: Mangala EOR Well Downhole Equipment Specification

ESP Electrical System Architecture

The ESP power system comprised of having a 33KV/50 Hz overhead line (OHL) system feeding power to various
well pads in Mangala and Thumbli. Two types of power distribution schemes were considered:
6 SPE 163117

a. One large step down transformer feeding power to multiple VFD at a well pad.
b. Small dedicated step down transformers feeding power to individual VFD.

Because of high lead time for a single large step down cum phase shift transformer and considering the flexibility
of operations and maintenance, it was decided to have dedicated step down cum phase shift transformer for the
two initial well pads viz. Thumbli water well pad and Mangala pad-12. The VFD selected for these two well pads
were 24 pulse systems to meet IEEE-519 requirement.

For the remaining well pads, after conducting a detailed cost analysis, the electrical system architecture was
revised to having a 33 KV/50 Hz packaged substation at each ESP well pad capable of stepping down the voltage
to 480 V/50 Hz and feeding power to each individual phase shift transformer which in turn fed power to the VFD. A
detailed electrical load analysis was conducted to understand the effect of using 6/12/24 pulse VFD system and it
was suggested that both 12 and 24 pulse VFD system are likely to comply with VTHD and TDD limits specified in
IEEE 519 for all the production cases. Hence, it was decided to use 12 pulse VFDs instead of the earlier choice of
24 pulse VFD system.

The overall electrical system architecture for two initial well pads and for subsequent ESP pads is shown in Figure
08 and 09.

ESP Control System Architecture

ESP control architecture was developed in the Yokogawa ICSS (Integrated Control and Safety System) for safe
and remote operation of the Mangala and Thumbli ESPs. (Figure 10) The basic philosophy of the control
architecture was as follows:

1. ESP operation to be made possible from ‘local’ i.e. at the VFD panel and from ‘remote’ i.e. from the
Mangala Processing Terminal (MPT) control room.
2. VFD to remain responsible for the protection of ESP downhole system irrespective of local or remote mode
of ESP operation. All downhole ESP and surface VFD related trips to be configured within the VFD.
3. ICSS to remain responsible for the overall process control and safety. ICSS to provide start permissive for
ESP to be started irrespective of local or remote mode of ESP operation. ICSS to initiate ESP trip in case
of unhealthy surface process parameters.
4. Bump less transfer of VFD control from local to remote mode and vice versa to ensure continuity of ESP
operations in the two possible modes of operation.
5. Provision of stop facility at the VFD as well as within the ICSS to stop VFD irrespective of VFD being in
local/ remote mode. ESP start to happen from local VFD panel only if the VFD is in local mode, similarly
ESP start to happen from remote MPT panel only if the VFD is set in remote mode. In any mode, VFD to
start only if the ICSS has provided start permissive to the VFD.
6. ESP frequency control possible from MPT if the VFD is set in remote mode.
7. Provision of frequency clamps at both VFD and at the ICSS to prevent accidental ESP operation beyond
the recommended range of operation.
8. A delay timer to be configured with the VFD to prevent accidental restart of ESP during the backspin
period. Time left for ESP restart to be relayed to MPT Human Machine Interface (HMI) over MODBUS®.
9. All ESP related downhole parameters and VFD related surface parameters to be communicated over
MODBUS® protocol and displayed and trended at the MPT HMI for ease of operation.
10. Alarms related to critical ESP downhole parameters and VFD related surface parameters to be configured
in the ICSS to initiate operator intervention if required.

With these features in mind an operator interface was developed for remote operations and monitoring of ESP
wells. (Figure 11)

Tornado Plotting

The importance of continuous real time monitoring of the operating point to ensure that the ESPs operate within
the recommended operating envelope was recognized at the beginning of the engineering stage. To that end,
pump performance envelopes at various frequencies were generated using affinity laws.
SPE 163117 7

To calculate total dynamic head (TDH) in ft of output per one stage based on a given flow rate Q (bpd) in BPD at
60 Hz:

TDH = C0 + C1*Q + C2*Q^2 + C3*Q^3 + C4*Q^4 +C5*Q^5 ……………………… (1)

Where:

C0, C1, C2, C3, C4 and C5 are pump coefficients obtained from the pump manufacturer and are characteristic of a
given pump.

To obtain TDH values for a multi-stage pump, above calculated TDH values must be multiplied by number of
stages in pump.

To convert a TDH value from 60 Hz standard value to another frequency, for one stage, the affinity law for
centrifugal pump head must be used:

TDH (new operating frequency) = TDH (60 Hz) *[(new operating frequency) / (60)] ^2 …………………… (2)

The vendor pump performance curve also provides the minimum flow, maximum flow and Best Efficiency Point
(BEP) flow for the pump type at 60 Hz. To convert a minimum flow, maximum flow, or BEP flow from the 60 Hz
standard values to another operating frequency, the affinity law for centrifugal pump flow must be used.

Q (new operating frequency) = Q (60 Hz) *[(new operating frequency) / (60)] ……………………… (3)

a. Water Wells

For Thumbli water wells the following input parameters were available:

1. Surface Flow Rate: Measured at the metering orifice.

2. Well head pressure: Measured by the pressure transmitter
3. Well Head temperature: Measured at the temperature transmitter
4. Pump Intake Pressure: Measured by the down hole sensor
5. Pump Intake temperature: Measured at the down hole sensor
6. Pump frequency: Measured at the VFD

Since the pump setting depth in feet (D) was known, the pump discharge pressure (Pd) was calculated as

Pd = Ps+ 0.052*ρw *D+14.73+ Frictional Losses ……………………… (4)

Where:

Ps: Surface Pressure in psig

ρw: Density of water in ppg (Default: 8.33 ppg for water wells)

A lookup table was created within the ICSS to calculate the frictional pressure losses for various surface flow rates
using standard pressure loss calculation equations.

TDH = (Pd-Pi)/ ρw ……………………… (5)

Where:

Pi: Pump Intake Pressure in psia

It was decided to plot the operating point at the pump discharge for which the surface flow rate was converted to
down hole flow rate using the formula:

Qd= Qs* Bw ……………………… (6)

Where:
Qs: Surface Measured Flow rate in bpd
Bw: Formation Water Volume Factor (at the pump discharge pressure Pd obtained from the PVT analysis of the
downhole water well sample)
8 SPE 163117

A lookup table of Bw against several possible discharge pressures was fed into the ICSS to calculate Bw against
the calculated discharge pressure

The operating point was plotted using two independent methods:

1. Plotting TDH against the operating frequency

2. Plotting Downhole flow against the operating frequency.

A good match for the two operating points was observed during the FAT and subsequently during the operations
phase. (Figure 12)

b. Oil Wells

For Mangala oil wells the following input parameters were available:

1. Surface Flow Rate: Measured at the MPFM, occasionally available

2. Well head pressure: Measured by the pressure transmitter
3. Well Head temperature: measured at the temperature transmitter
4. Pump Intake Pressure: measured by the down hole sensor
5. Pump Discharge Pressure: measured by the down hole sensor
6. Pump Intake temperature: Measured at the down hole sensor
7. Pump frequency: Measured at the VFD

For the purpose of calculation, it was assumed that the pump discharge temperature is same as the pump intake
temperature.

It was recognized early on that the plotting of operating point for oil well ESPs will be difficult due to:

1. Absence of continuous surface flow metering for each ESP well, since there was only one MPFM per well
pad which catered to the metering requirement of multiple wells flowing from a well pad.
2. Challenges in converting the surface flow rate to downhole rate.
3. High oil viscosity and expected emulsion viscosity which would require viscosity correction of the pump
tornado curve.

Nevertheless, a step by step approach was taken to solve the seemingly complex problem:

Step1: Generation of pump tornado curve

For various possible mixture viscosity µm values, head (Hm) and flow modifiers (Qm) were obtained for the ESP type
from the vendor published information.

Actual target pump rate (Q) at any operating frequency (F) is assumed.

Calculate adjusted rate (Qa) as follows:

a. Qa = Q * (60 / F) / Qm ……………………… (7)

Calculate unadjusted head using adjusted rate (Qa) as follows:

a. H = H0 + H1 * Qa + H2 * Qa^2 + H3 * Qa^3 + H4 * Qa^4 + H5 * Qa^5 ……………………….(8)

Calculate head (Ha) adjusted for frequency and viscosity effects as follows:
a. Ha = H * Hm * (F/60)^2 ……………………….(9)

The point on the rate-head curve is Q, Ha.

Using the modifiers, the minimum, maximum and the BEP flow values at different operating frequencies ranging
from 35 to 60 Hz was modified and the pump tornado curve was created in the ICSS.

Step 2: Calculation of downhole discharge flow rate

SPE 163117 9

Qdischarge = Qoil (sc) (Bo) + Qwater (sc) (Bw) + Qoil (sc) [(Qgas (sc)/ Qoil (sc)) – Rs] Bg ………………………. (10)
Qdischarge = Volumetric Flow rate at pump discharge in bpd
Qoil (sc) = Oil Flow rate at standard conditions in bpd measured from MPFM
Qwater (sc) = Water Flow rate at standard conditions in bpd measured from MPFM
Qgas (sc) = Gas Flow rate at standard conditions in scf measured from MPFM
Bo = Oil Formation Volume factor at discharge pressure from Look Up table
Bw = Water Formation Volume factor at discharge pressure from Look Up table
Bg = Gas Formation Volume factor at discharge pressure from Look Up table
Rs = Solution gas Oil ratio in scf/bbl, at discharge pressure from Look Up table

In case [(Qgas (sc)/ Qoil (sc)) – Rs] < 0, the ICSS program should read it as zero.

Note: All formation volume factors and solution GOR to be used from look up tables built within the ICSS from PVT
report assuming pump discharge temperature is same as reservoir temperature of 149 deg F.

Step 3: Calculation of downhole mixture density

M= [ρgas *0.0765* (Qgas (sc)/ Qoil (sc)) + (141.5/ (API+131.5)*350.28]* Qoil (sc) +42*ρwater* Qwater (sc) …………………. (11)
ρdischarge = M /Qdischarge …………………. (12)
Where:
M: Mass Flow rate (lb/day)
ρdischarge: Discharge Density (lb/bbl)
ρgas: SG of gas, to be User Input (Default = 0.73)
API: To be User Input (Default = 29.2)
ρwater: Formation Water Density: To Be User Input (Default = 8.40 ppg)

Step 4: Calculation of TDH

TDH= (Pd-Pi)*2.31/ ρdischarge …………………. (13)

Once the TDH and downhole flow rate can be plotted, it was easy to plot the operating point using the following
methods:

1. Plotting TDH against the operating frequency

2. Plotting downhole flow rate against the operating frequency

As stated earlier, it was recognized that the downhole flow rate calculations will not be available at all times
because of non availability of MPFM continuously for one single ESP well. Hence, it was decided that the mixture
density will be calculated based on the last available well test and this parameter will be assumed constant until
the next test. This was not considered an unfair assumption given the importance of horizontal wells and these
wells were expected to be tested at least once a week. Moreover, the fluid behavior like GOR, water cut etc. was
not expected to change significantly during a week. The ICSS methodology was identified with the following
limitations:

1. Does not account for gas slippage for calculating down hole flow rate and TDH
2. Uses mixture density at pump discharge instead of calculating mixture density at each stage inlet and then
calculating overall head developed by the pump by summation of TDH at every pump stage
3. Tornado curve correction for mixture viscosity when the mixture viscosity is difficult to measure in real time.
4. Assumes same mixture density between two well tests

Given these limitations in the ICSS methodology of operating point calculations, an exercise was undertaken to
validate the operating points calculated in the ICSS system against the operating point calculated in the Autograph
PC. Against possible surface liquid production rates, produced GOR and water cut, pump discharge flow rate
calculated using ICSS methodology and Autograph PC® were compared.(Figure:12) Also, the TDH calculated by
the ICSS methodology was compared with the TDH calculated by Autograph PC® under various possible
production scenarios. (Figure: 13)

It can be seen that the error percentage of the pump discharge rate and TDH were less than 5% in the worst case
GOR scenario of 500 scf/bbl (Solution GOR: 192 scf/bbl) once the water cut was beyond 40%. It was predicted
10 SPE 163117

that ESPs would be mostly required once the water cut exceeds 40% and hence the ICSS methodology was able
to reasonably ascertain the operating point.

Operator Training

One of the key aspects of the RJ ESP project was the near absence of previous ESP operating experience in the
operator company’s field operations and office based engineering staff. A series of training programs were
conducted ahead of first ESP installation and commissioning to disseminate the required technical knowhow of
ESPs. This also involved getting operators to operate dummy HMI panels and simulate ESP operations. Another
important aspect was to sensitize the senior management about ESP operations. These training interventions
proved useful and have been one of the important factors in the successful operations of the ESPs in Rajasthan
fields. Frequent refresher training courses have been conducted to keep providing the plant operations team with
necessary inputs; this has also helped as the plant operators have been able to clear various doubts on ESP
related operations after getting hands on experience of operating ESPs as the ESP systems mature and the
project moves from one stage to another.

ESP Installation Experiences

After extensive preparations, including a complete-the-well-on-paper (CWOP) exercise, the first ESP completion
was run in a water well during the month of November’ 2009. This was a relatively simple installation and went
smoothly without any problems. Nevertheless, it gave the necessary confidence to the installation crew and the
office based team of engineers ahead of the critical Horizontal well ESP completions which involved running 1. 5”
heater string and a 2 X 3/8” chemical injection line flat pack in addition to the #2 AWG cable and the 4.5” main
completion string with an eccentric tubing hanger. During the month of April 2010, the first two oil well ESP
completions were run, of these the first completion was landed successfully, however as the second completion
was about to be landed on the tubing hanger, a loss of electrical integrity was observed in the ESP cable,
subsequently problems were also faced in retrieving the ESP completion due to problems with unsetting the ESP
packer. Finally the packer was released using over-pull. After pulling out, it was found that the cable failure
happened at two locations, one immediately below the packer in the MLE against seal section and another in the
main ESP cable below the tubing hanger above the last splice. Cable bands were extensively used to support the
MLE at various locations in the completion section below the packer and at the splices above the packer and most
of them were found to be left in hole. A lot of effort was made in fishing out these bands from the top of reservoir
control valve.

A detailed failure analysis was conducted and the following recommendations were implemented to avoid similar
failures in future:

1. Avoid use of cable bands in all horizontal well ESP installations where the ESP assembly was designed to
be landed in deviation of 60o and above.
2. Design and procurement of higher standoff neck clamps to support the MLE and similarly higher standoff
main cable clamp to support the main cable, heater string and the chemical injection flat pack.
3. Problems associated with retrieving the ESP packer were also analyzed and it was recommended to set
the ESP packer using revised guidelines from the supplier. These involved pressurizing tubing and
annulus alternately in stages to achieve proper packer set.
4. Using additional pup joints and cross coupling clamps in the completion string at those points where the
ESP cable was found prone to slack or slumping, for example at the splices.
5. Use of additional pup joints and cross coupling clamps immediately below the tubing hanger to provide
adequate support and centralization to main ESP cable during the critical hanger landing operation.
6. Careful space out of ESP splices to avoid splice against a tubing coupling.

With the revised procedures in place and the procurement of high offset clamps, the ESP installation activities
resumed again and a campaign was started to install ESPs in all remaining water wells and the high PI horizontal
oil wells. All these planned ESP completions went smoothly without any issues reported during the installation
process,

Water Well ESP Commissioning and Operations Experience

SPE 163117 11

The first water well ESP was commissioned in the month of July’2010 and provided significant learning to the ESP
project team. The productivity index of the well was extremely high at 2700 stb/d/psi, compared to the initial
estimate of 750 stb/d/psi. This was attributed to better drilling and natural gravel pack development practices. With
the better than expected PI and absence of any back pressure control mechanism at the water well pad, there was
a significant risk of pump operating in the up thrust region. As a short term measure, to commence ESP
commissioning operations, it was decided to install a suitably sized restriction orifice (RO) in a straight flanged pipe
section. Various calculations were run in Autograph PC® to predict the performance of the pump at different
frequencies and against different wellhead pressures expected due to the presence of fixed RO in the line. A step
wise pre-commissioning exercise was completed to check the electrical and control functionality of the first water
well ESP system. The water well pad is ~ 20 Km away from the Mangala Processing Terminal- Central Control
Room (MPT- CCR), in order to facilitate quick response by the commissioning team a standalone local HMI
configured to control operations for water well pad. It was initially observed that the remote VFD operation was not
possible because of a circular logic built in the start permissive, wherein the VFD was expecting an all clear signal
from the ICSS to generate a drive “ready to run” signal and at the same time the ICSS expected a drive “ready to
run” signal to give an all clear signal to the VFD. The circular logic built in the system was modified and then local
and remote operations functionality of the VFD from the local VFD panel, stand alone HMI panel and MPT- CCR
was successfully tested under no load conditions. All the process related cause and effects were sequentially
tested. The ESP was started at 35 Hz frequency using standard operating procedure, however significant
difference in phase current was observed and this caused the ESP to trip frequently because of higher the
permissible current imbalance between the phases. This was attributed to improper functioning of FPWM and it
was decided to complete commissioning in the ESP mode of VFD operations wherein balanced current was
observed in all outgoing phases from the VFD.

After operating the ESP at the well for 7 days using the local HMI, the ESP operations were handed over to the
MPT- CCR for conducting regular remote operations and monitoring from MPT. In subsequent months, two more
water well ESPs were commissioned which again had similar problems in the FPWM mode. A detailed analysis of
the problem was conducted which led to the following outcomes:

1. Problems within the installed version of the VFD software which caused improper convertor firing and
hence the current imbalance when the drive was put on load.
2. The FAT procedure involved testing the 24 pulse or 12 pulse VFD in 6 pulse mode and hence the problem
with FPWM went unnoticed during the factory tests.

To avoid repeating the same problems in future wells, particularly in the more critical horizontal oil well ESP
commissioning operations, the following plan of action was identified:

1. Mobilization of 100 % resistive load bank at all oil ESP well pads which have topside systems installed and
ready for ESP commissioning. The procedure involves running the VFD on load for up to 24 hours using
the resistive load bank and troubleshoot FPWM.
2. Change in FAT procedure, wherein all the remaining VFDs were to be subjected to mandatory full load
testing in FPWM mode prior to FAT acceptance.

After the commissioning of the first three water well ESPs, RO from each line was removed and back pressure
controller for each well was installed to operate the ESPs on a long term continuous basis. Management of multi-
disciplinary team responsible for complex ESP operations involving significant level of electrical, petroleum,
process and instrumentation engineering was identified as a key challenge area. It was decided that better
communication mechanisms such as pre-commissioning meetings, discussions on lessons learnt from previous
commissioning experiences with greater participation from the multi- disciplinary personnel actually involved in the
commissioning can immensely benefit subsequent ESP well commissioning.

Oil Well ESP Commissioning and Operations Experience

The first oil well ESP was commissioned in Dec’ 2010. Having successfully tested the remaining functionality built
in the ICSS, one of the key issues identified was correct tornado plotting and determination of the operating point
in actual field conditions. After starting the ESP, a wide variation in the two operating points was observed.
(Figure: 14) It was suspected that either the well was producing at a higher GOR or the downhole pressure
sensors were not reading the correct values. To resolve between these possibilities, calibration of the surface
MPFM was conducted and PVT data input to the MPFM was checked, calibration of down hole gauges was
conducted by running slick line memory gauges. The slick line memory gauge data were in conformance with the
downhole ESP gauges. A better match of vertical lift performance (VLP) of the well, between the pump discharge
12 SPE 163117

and tubing head, at a higher GOR was also obtained by running PROSPER®. In a subsequent ESP well, which
had moderate GOR, a good match of the MPFM measured flow rates and a PROSPER® calculated VLP was
obtained, thus establishing that the MPFM readings were correct. A good match of the two ESP operating points
was found for this well, thus establishing that ICSS calculations were able to ascertain the pump operating point
correctly provided the MPFM readings were reasonably accurate. (Figure: 15) Since then, a number of well tests
conducted on self flowing and jet pumped wells with high GOR and high water cut, spurious test reports have
occurred which could not be satisfactorily matched in PROSPER®. It was suspected that the MPFM was unable to
measure the flow rates especially in high GOR and high water cut cases. Current effort is focused on migrating
from the current MPFM technology to another technology which can provide better results in respect of GOR and
water cut for viscous reservoir fluids such as Mangala crude oil.

In some of horizontal oil wells producing at high water cut, even after the ESP completion was landed, the topside
equipment installation and commissioning took longer than expected and it was decided to produce these wells on
jet pumps in the interim. While doing so, it was recognized that the hot water jet pumping pressure cycles can be
detrimental to ESP cable and procedures were put in place for gradual increase to available jet pumping pressure
of 2500 psi in stages and ramp down the power fluid pressure in stages. However, during ESP commissioning, one
horizontal well which was on jet pumping failed immediately after the ESP was started. A Time Domain
Reflectometer (TDR) survey was carried out and the ESP cable fault was suspected to be located immediately
below the tubing hanger. During pull out of the system, ESP cable damage was observed 4 meters below the
tubing hanger at the cable splice with an additional fault located above the packer in the packer penetrator.
Moreover severe general corrosion was seen all over the ESP galvanized steel armor. (Figure: 16 & 17) Detailed
analysis revealed that, ESP cable was damaged due to decompression damage of cable insulation which was
compounded by heavy armor corrosion. The severe general corrosion in the galvanized steel cable armor was
attributed to Zinc acting as a sacrificing material under the corrosive hot water environment with high rate of
corrosion because of stagnant conditions. Although, the hot water was chemically inhibited, it was suspected that
the inhibition was not effective under stagnant conditions. As an immediate measure hot water circulation has been
stopped in all ESP wells and the annulus hot water fluid will be replaced with conventional packer fluid comprising
fresh water with appropriate dosage of oxygen scavenger, corrosion inhibitor and biocide. As a long term measure
it is also planned to change cable armor metallurgy to Monel and avoid jet pumping operations in any viable ESP
well.

ESP operations in RJ have also been marred by a few step up, step down and phase shift transformer and VFD
failures. At the time of writing this paper, detailed investigation to ascertain the reason for these failures is under
progress. As an immediate measure, a preventive maintenance procedure has been developed and is being
rigorously followed.

Recent PI Reduction in Oil Wells

Almost all the horizontal ESP wells in Mangala have seen decline in well productivity at high withdrawal rates and
especially after the onset of water production. (Figure: 18) During Workover of the well with ESP cable failure a
PLT was run to understand the inflow profile of the horizontal section and strong gamma ray spikes were seen,
which are normally associated with barium and strontium sulfate.

A threefold damage mechanism is suspected to be in place in Mangala horizontal wells:

1. Damage due fines migration- Supported by the fact that some of the negligible water cut wells had seen
drastic PI decline at high withdrawal rates. (Figure: 19)
2. Damage due to carbonate scaling: Supported by the fact that Mangala crude has high CO2 content and
Thumbli injection water is rich in calcium ions. A comparative chemical analysis of Mangala formation
water and Thumbli source water is shown in Figure 20.
3. Damage due to sulfate scaling: Supported by the GR spikes seen in PLT logs in various wells and barium/
strontium scaling found in jet pumps in deviated producers in Mangala. (Figure: 21)

An HCl acid stimulation program was implemented in all horizontal wells and there were significant PI
improvements which tapered out in subsequent months (Figure: 18). In one well on a pilot basis, to understand
whether the damage was due to screen plugging a clay inhibited water bullhead program was implemented and
this also saw a significant improvement, but less than the PI improvement due to HCl treatment. An HCL/DTPA
stimulation program is currently under progress to regain horizontal well productivity. Meanwhile laboratory work is
under progress for selection of suitable scale inhibitor and fines fixer for Mangala wells; treatments are planned in
the near future.
SPE 163117 13

Conclusions

Installing and operating a successful ESP system is the product of several activities, all of which must be
performed to a consistent, high standard.

1. Water well and oil well ESPs and well completions have been designed and specified to a high standard.

2. Installation requires careful planning and preparation (e.g. CWOP). Be ready to respond to early issues,
(eg cable management and improved cable clamp location)

3. The ESP control & protection system is important to success. The ‘Tornado’ HMI graphics are a very
useful aid to maintaining each ESP in a healthy operating regime, consistent with achieving long ESP life.

4. Training is essential. ESP operators need to understand the key fundamentals to keep an ESP healthy,
identify problems, assess the problem and decide a course of action. Experienced / expert support and
advice to resolve problems has been useful.

5. Commissioning, ESP start-up and early operations required focus (and experienced support) to ensure no
early ESP failures.

6. Chemical scaling, corrosion and fines movement have all occurred and all have been more severe than
anticipated. However development planning and hardware specifications had considered that possibility
and remedial measures are proving successful.

Overall, RJ ESP Project has been successful in delivering and sustaining plateau production from Mangala field
and in meeting the water production requirement from Thumbli source water field. Despite being a high stake
green field development in a geographic location with limited ESP infrastructure the project team has been able to
deliver a satisfactory ESP performance. A lot of learning has been obtained during the last four years of
engineering design and field execution which will immensely help the team to deliver future ESP projects in the
satellite fields of Mangala.

Nomenclature

AWG American Wire Gauge

BEP Best Efficiency Point
BGL Below Ground Level
CRC Corrosion Resistant Coating
CWOP Complete Well on Paper
EOR Enhanced Oil Recovery
ESP Electrical Submersible Pump
FAT Factory Acceptance Test
FPWM Filter Pulse Width Modifier
GOR Gas Oil Ratio
HMI Human Machine Interface
HP Horse power
ICD Inflow Control Device
ICSS Integrated Control & Safety System
IEEE Institute of Electrical & Electronics Engineers
MLE Motor Lead Extension
MPFM Multi Phase Flow Meter
MPT Mangala Processing Terminal
MSL Mean Sea Level
NAB Nickel Aluminum Bronze
OHL Over Head Line
OWC Oil Water Contact
PI Productivity Index
RO Restriction Orifice
14 SPE 163117

SAS Stand Alone Screen

STOIIP Stock Tank Oil Initially In Place
TDH Total Dynamic Head
TDR Time Domain Reflectometer
TDS Total Dissolved Solids
VFD Variable Frequency Drive
WAT Wax Appearance Temperature

Acknowledgement

The authors would like to thank Cairn India Limited and ONGC, the JV partners of the RJ-ON-90/1 block and
Triangle Engineering and Baker Hughes for their permission to publish this work. We would also like to thank Tim
Mc Kenzie, Divyendu Tripathi, Piyush Kumar, Lourdu Kanagam Soundarajan, Girish Kant Gupta (Cairn India
Limited), Vijay Nambiar, Dilip Patel, Prasannan (Baker Hughes), Amit Chauhan, Sridharan Rangachari (Yokogawa
India Limited) for their significant contributions at various stages of this project.

References

1. The technology of artificial lift methods, Kermit E. Brown etal.

2. SPE 108381-MS: Cliff Head intelligent completion with coiled tubing deployed ESP- Increased production,
reduced life cycle cost, Chris Way etal.
3. SPE 128003-PA: From Operations to Desktop Analysis to Field Implementation: Well and ESP
Optimization for Production Enhancement in the Cliff Head Field, Z.F. Dholkawala etal.
4. SPE 151134- MS: Innovative well completions meet the challenges of producing high wax crude in a giant
onshore field in India, Vipin Kumar etal.
SPE 163117 15

Figure1: RJ-ON-90/1 Block

16 SPE 163117

Figure 2: Structural Model of Mangala Field with Top Fatehgarh Surface and OWC

Figure 3: Overview of Mangala field and Thumbli Aquifer

SPE 163117 17

Figure 4: Mangala Horizontal Well with ESP completion

18 SPE 163117

Figure 5: Thumbli Water Well with ESP completion

SPE 163117 19

Figure 6: Mangala Expected Emulsion Viscosity under Typical ESP Shear Conditions

Figure 7: Mangala EOR Well with ESP completion

20 SPE 163117

Figure 8: Electrical Power Scheme at Thumbli Water Well Pad and Mangala Pad-12

Figure 9: Electrical Power Scheme at Remaining ESP Pads

SPE 163117 21

Figure 10: ESP Control System Overview

Figure 11: Remote ESP Operations and Monitoring Interface

22 SPE 163117

Figure 11: Water Well Tornado Plot

Figure 12: Pump Discharge Rate Error in ICSS Methodology

SPE 163117 23

Figure 13: TDH Error in ICSS Methodology

Figure 14: Poor Match in Operating Points for Oil Well with Incorrect MPFM Reading
24 SPE 163117

Figure 15: Good Match in Operating Points for Oil Well with Correct MPFM Reading

Figure 16: Decompression Damage of ESP Cable due to Jet Pumping

SPE 163117 25

Figure 17: Severe Corrosion of ESP Cable due to Incompatibility with Annulus Hot Water

Figure 18: Decline in PI in a Horizontal Well with Water Production

Figure 19: Decline in PI in a Horizontal Well with Negligible Water Production

26 SPE 163117

Figure 20: Comparative Chemical Analysis of Thumbli Water and Mangala Formation Water

Figure 21: Scaling in Horizontal Wells with High GR Activity Observed During Production Logging