S.G. Noonan, ConocoPhillips; W. Klaczek, K. Piers, C-FER Technologies; L. Seince, PCM; S. Jahn, Kudu Industries
Copyright 2008, SPE/PS/CHOA International Thermal Operations and Heavy Oil Symposium
This paper was prepared for presentation at the 2008 SPE International Thermal Operations and Heavy Oil Symposium held in Calgary, Alberta, Canada, 20–23 October 2008.
This paper was selected for presentation by an SPE/PS/CHOA Program Committee following review of information contained in a proposal submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers, the Petroleum Society of Canada, or the Canadian Heavy Oil Association and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the SPE/PS/CHOA, its officers, or members. Papers presented at SPE, PS, and CHOA meetings are subject to publication review by Editorial Committees of the SPE and PS. Electronic reproduction, distribution or storage of any part of this paper for commercial purposes without the written consent of the SPE or PS is prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300 words; illustrations may not be copied. The proposal must contain conspicuous acknowledgement of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435 and Editor, Journal of Canadian Petroleum Technology, Petroleum Society of Canada, Suite 425, 500 - 5th Avenue S.W., Calgary, AB, Canada T2P 3L5, fax 01-403-262-4792.
Abstract
ConocoPhillips has been on a quest to find a high volume artificial lift system that will operate reliably in a 250°C (482o
F) downhole environment, which exists in certain SAGD applications. This presented two problems: 1) there were no commercially available technologies for such a high temperature; and 2) there were no facilities capable of testing these systems.
This paper describes the complexity of building and operating a high temperature flow loop rated for 250°C, and the lessons learned while upgrading an existing flow loop, from the initial design through the final commissioning phases. The paper also describes the issues encountered with the first artificial lift system tested at 250°C, which was a metallic stator
progressing cavity pump system, rated for 1100 m3/d (6919 bpd) at 500 rpm.
In the end, the test program not only served to validate and define the pump’s performance, but also provided valuable lessons on the completion configuration and operational procedures.
Introduction
The ConocoPhillips technology group was tasked to find, select, and further develop artificial lift technology with the
capability of handling fluid rates up to 1000 m3/d at 250°C (6290 bpd at 482oF) downhole conditions. The goal was not to
just find and validate a single system, but to qualify several lift systems, in order to provide the production engineers with a “toolbox” of solutions.
This challenge was divided and approached as two different projects: 1) find, select and further develop potential lift systems with the needed volumetric capability; and 2) validate these systems through high temperature testing. The latter was considered to be the bigger challenge of the two. ConocoPhillips did not operate any fields with downhole temperatures close to 250°C, so validation via field trial was not possible. A more controlled test facility (whether a well or flow loop) was also preferred, so that a comprehensive suite of performance curves could be collected to define the full operating envelope for each lift candidate. A test facility which was not associated with a specific pump vendor was also preferred, to avoid the legal and confidentiality issues with testing third party equipment.
ConocoPhillips decided that an existing high temperature flow loop located at C-FER Technologies Ltd, in Edmonton, Alberta, Canada was the best option for the artificial lift validation testing. The loop had been built as part of a Joint Industry
Project (JIP)1 in 2004, but needed to be upgraded to allow for testing at 250°C. This was a costly endeavor, and
ConocoPhillips and C-FER contacted other Canadian SAGD operators to see if the upgrade could be completed as a JIP, thus sharing the capital cost among several interested parties. However, no other operators were interested in upgrading the loop at the time, so ConocoPhillips proceeded to fund the project entirely.
While the flow loop upgrades were in progress, the search for artificial lift systems also began. The search confirmed that there were few lift systems commercially available in mid-2007 that had both the required volumetric capacity and high temperature rating. Therefore, while ConocoPhillips is currently developing several lift systems specifically to meet this criteria, it was decided that it was best to first try to validate systems already being used by other SAGD operators, albeit at lower downhole temperatures. The first system purchased for the test program was the metallic stator Progressing Cavity Pump (PCP) manufactured by PCM and sold in Canada by Kudu Industries.
Upgrading the Existing High Temperature Flow Loop
As mentioned above, C-FER had already designed and built a high temperature flow loop as part of a multi-company JIP, in which ConocoPhillips was a participant During this project, various artificial lift systems were tested, at fluid rates up to
900 m3/day at 200°C (5660 bpd at 392oF), under both high and low intake pressure conditions, and with varying fluid
compositions (i.e. oil, water, oil and water mixtures, and up to 20% free gas fractions at the intake). The basic schematic of the flow loop is provided in Figure 1.
Figure 1: Basic Schematic of the C-FER Flow Loop
Design, Procurement and Construction
Similar to the original design, the upgraded flow loop had to accommodate a wide variety of downhole pumping systems and test conditions. The goal was to retain the same operating capabilities in terms of flow rates and fluid types but to upgrade the temperature capability to 250°C. As most tests were to be conducted above the steam saturation pressure at the test temperature, the pressure rating of several components also needed to be increased.
The procurement of suitable high temperature and pressure equipment for the upgrade was often quite difficult. During the course of the loop upgrade a large portion of the pressure piping and much of the instrumentation was replaced. Examples of the components that were replaced or added include: high pressure and temperature Coriolis mass flow meters, control valves, temperature and level transmitters, a circulation pump, an air compressor, and several new safety related items (e.g. steel blast doors, double-walled storage tanks, personal protective equipment, video cameras, etc.).
Finding a Suitable High Temperature Test Fluid
Finding suitable high temperature test fluid was identified at an early stage as one of the most significant technical challenges for the test program. Safety is always a top consideration for all the parties involved. Since the flow loop intake pressure is controlled using high pressure air (i.e. it is a modified open system design), and additionally some tests are performed with a fluid/air mixture, the policy is that all test fluids must have a closed-cup flashpoint higher than the maximum test
temperature. Locating suitable test fluid with this characteristic for use at 250oC was extremely difficult. Eventually, two
alternatives were identified, but, it was later learned that one of the fluids was incompatible with water at high temperatures (as described later in the paper).
9-5/8” Main Casing
Coriolis Flow Meter & Back Pressure Control Coriolis Flow Meter
Heat Exchanger Heater Transfer Pump Oil Storage Tank Downhole Pump to be Tested Inclined Separator (Pressure & Level control)
Wellhead
Air Compressor Coriolis Flow Meter
& Pressure Control
Emulsion Storage Tank
Main Separator (Pressure & Level control)
System Commissioning and Verification
System commissioning and verification was conducted after the loop upgrades, to verify the capabilities of the system. Initial commissioning attempts resulted in additional insulation being added to the loop, and an additional temperature measurement being added inside of the heater. Final commissioning was completed with water, after circulating fluid at 250°C for an extended period of time to check and fine-tune the temperature and pressure control instrumentation and the overall operation of the flow loop.
Final Upgraded Capabilities of the Flow Loop
A basic summary of the final flow loop capabilities, following the high temperature upgrade, are provided below:
o The physical setup of the flow loop is very similar to the original design. The main casing section consists of approximately 26.5 m (87 ft) of 9-5/8” casing that is orientated at 87° (i.e. nearly horizontal to simulate SAGD installations), allowing for pumping sytems up to approximately 24 m (80 ft) to be tested.
o Pump intake temperature can be controlled between 30°C and 250°C (86°F and 482°F) with an electric heater and power controller, and with a water-chilled heat exchanger to provide cooling.
o Pump intake pressure can be controlled between approximately 0.1 to 4.5 MPa (15 psi and 650 psi) by regulating the pressure in the separator using a dedicated air compressor system.
o Pump discharge pressure as high as 12.4 MPa (1,795 psi) can be achieved without vibrational isolation, or 8.3 MPa (1,200 psi) with vibrational isolation between the casing section and the remainder of the flow loop.
o Volumetric flow rates can be controlled between 25 m3/d (157 bpd) and 800 m3/d (5,000 bpd), with either oil, water,
or an oil/water mixture while at high temperature. Higher flow rates can possibly be achieved at lower viscosity conditions.
o Maximum air injection rates of approximately 119 sm3/hr at 4.5 MPa (70 scfm at 650 psi) of system intake pressure
are possible.
o Real-time measurements of pressure, temperature, flow rate, pump speed, vibration, etc., are obtained using a custom designed data acquisition and control system that directly interfaces with instrumentation and system components within the flow loop.
Some images of the upgraded flow loop are provided in Figure 2.
Figure 2: Images of the Upgraded C-FER Flow Loop
9-5/8” Casing Wellhead & Flow Tee Annular Separator High Pressure Discharge Section Strainer Coriolis Meter Fluid Flow Direction
Testing the Metallic Stator PCP – Round One
While the metallic PCP is not entirely new to the SAGD industry, it has gone through design changes and product expansions
as knowledge is gained from the various field trials2. The pump purchased for this test was the first of its size (i.e. the largest
capacity offered: 1100 m3/d at 500 RPM with a pressure rating of 750 m lift) and had design modifications (V2) to correct
failure modes observed with earlier versions (V1) installed at Cold Lake and Joslyn3
o Stator connection to the external envelope reinforced to minimize displacement and improve energy transfer to the external envelop
o Pump installation in tangent section to decrease stress on the equipment
o No-turn tool / anti-vibration sub used so that intrinsic unbalance is absorbed through the casing o Well head pressure set high enough to keep rod string in tension to eliminate rod buckling o Additives in coating to improve life span
o Improved quality control for better alignment and tighter rotor stator fit
These improvements were made possible by a complete detailed failure analysis (macroscopic and microscopic), dynamic modeling of the pump and system using Fast Fourier Transform analysis (FFT) as well as numerical simulation. Calculation of dynamic behavior was validated with test bench measurements.
Pump Specifications and the Experimental Setup – Round One
The pump is comprised of two main components: the stator and the rotor. The stator has a dual helical profile which is fully metallic and produced by hydroforming. The rotor has a single helical profile and rotates inside the stator with a positive clearance between the two. Both are specially coated for high temperature and wear resistance, however the rotor is designed to be the sacrificial element.
A technical summary of the downhole and surface assembly used during the first round of testing is provided below: o Metallic Stator: total length of 8.796 m (28’10”), overall outer diameter of 0.135 m (5.31”).
o Metallic Rotor: total length of 9.015 m (29’ 6”).
o Pump Specifications: 12 stages, 750 m (2,460 ft) of head capability, 1636 cc displacement, with a nominal capacity
of 2.35 m3/d/RPM (14.8 bpd/RPM).
o Completion details: 4-1/2” NUE pin on stator discharge, attached to 15.25 m (50 ft) of 4-1/2” production tubing to surface, swedged to a 3-1/2” EUE wellhead bonnet, through a #3000 Flow Tee, and connected to the flow loop. o Drive Completion details: rotor was attached via approximately 15.25 m (50 ft) of 1-1/2” rod with 1-1/8” pin
connections, comprised entirely of 1.82 m (6 ft) and 3 m (10 ft) pony rods, with 6 evenly spaced concentric spin- through centralizers located on the rod string. The spin-through centralizers were new prototypes that were to be tested at high temperature at the same time as the pumping system.
o Surface Drivehead: 74.5 kW (100 hp) electric drivehead at surface, controlled with a Variable Frequency Drive (VFD).
It was decided that no tag bar, special intake (e.g. tail pipe, bottom feeder), casing centralizer, or torque anchor would be included in this test configuration. The tag bar and additional intakes were not included to prevent restrictions or pressure drops at the pump intake, so that the true pump performance could be measured more accurately. Centralizers and torque anchors were omitted at the request of ConocoPhillips because they would make it more difficult to measure pump vibration due to eccentricity. In addition, ConocoPhillips would like to eliminate these items from field installations and needed to understand how the pump would perform without them.
Testing – Round One
The objective during the first round of testing was to simulate a cool, slow startup phase, and then increase the temperature
and pump speed to the peak expected field operating conditions of 250°C and 1000 m3
/d. To further simulate expected field conditions, a water cut of approximately 70% was maintained throughout the first round using synthetic high temperature gear oil and potable water. To gauge repeatability and determine whether pump performance had changed after testing at
250°C, a series of “repeat” tests at 200o
The original test matrix for the first round of testing is provided in Table 1.
70% Water Cut (+/- 5%)
Temperature 150 RPM 200 RPM 300 RPM 400 RPM
90°C Pump Curve
150°C Pump Curve Pump Curve PIP Reduction ALR Sensitivity 200°C
Pump Curve
PIP Reduction Pump Curve & Air PIP Reduction & Air
Pump Curve PIP Reduction
Pump Curve 220°C Pump Curve Pump Curve Pump Curve 250°C Pump PIP Reduction Curve
Shutdown/Restart Test
Pump Curve PIP Reduction
Pump Curve 200°C Pump Shutdown/Restart Test Curve Pump Curve Pump Curve 150°C Pump Curve Pump Curve ALR Sensitivity
Table 1: Test Matrix for Round One
The details of the various tests planned for the test matrix at the different levels of temperature and RPM are as follows: o “Pump Curve” refers to a standard pump performance test, where pump flow rate is monitored while varying the ΔP
across the pump and holding the other test conditions constant (i.e. temperature, intake pressure, water cut, etc.). o “PIP Reduction” refers to a test where the pump flow rate is monitored while the pump intake pressure is reduced,
and the ΔP across the pump and other test conditions are held constant.
o “ALR Sensitivity” refers to a test where the pump flow rate is monitored as the air injection rate (i.e. the air/liquid ratio) is increased, and the ΔP across the pump and other test conditions are held constant
o “& Air” refers to the same type of testing but with a constant amount of air injection included.
o “Shutdown/Restart Tests” were added to test operational conditions when re-starting the system at high temperature.
This test program commenced in November 2007, beginning with the lowest fluid temperature of 90°C (194o
F). Volumetric performance of the pumping system was generally as expected. However, a high level of system vibration was observed intermittently during testing at this temperature, at one point resulting in some minor equipment damage within the flow loop, and creating additional safety concerns for the upcoming higher temperature tests. It was unclear whether this vibration was originating from the metal-to-metal PCP, the rod string, or both. It was also originally expected that the static break-out torque of the metal pump would be minimal, as there was no interference fit between the rotor and stator (i.e. unlike an elastomeric PCP); however, a high amount of torque was observed at times.
After some results were collected at 90°C, testing at 150o
C began. At this temperature, some higher levels of system vibration were experienced. At this point, testing was temporarily halted pending an operational safety review, and some
commissioning activities at higher temperatures (up to 215oC using the test fluid, but with the pump not rotating) were
conducted. A further complication occurred after this high temperature commissioning, when unknown abrasive particulates were observed in the test fluid (i.e. retained in a strainer). It was not known at the time what the particles were, or if they would interfere with normal pump operation. Based on these factors, the decision was made to halt the test program until a better understanding of these problems could be gained, and the downhole equipment could be removed and inspected for signs of wear or imminent failure.
Following the pump removal from the loop, it was observed that the pumping system, the drive string, and spin-thru centralizers all appeared to be in excellent condition. However, it was found that the test oil, upon further analysis, had degraded due to incompatibility with water at high temperature, with the particulates being a by-product of that degradation. It was also hypothesized that the drive rod configuration (i.e. a relatively short length with six spin-thru centralizers) was too stiff to absorb all the pump eccentricity, resulting in significant forces against the flow loop tubing, and thus causing the pump and tubing string to move within the flow loop casing, and the flow loop to vibrate. The loop had been designed without rigid constraint between the casing and casing supports so that the casing could grow axially with thermal expansion. This had worked well in previous testing; however the loop had not been designed to deal with the large radial forces that were being experienced in the first round of testing.
Lessons Learned – Round One
Unfortunately, it was not possible to complete the entire test matrix, as the program was halted before high temperature testing was completed. However, in addition to some lower temperature results, a number of important observations were made regarding the operation of the pumping system and the experimental setup.
Some of the key lessons learned are summarized below:
1. In addition to the flashpoint and viscosity tests that were performed, the test oil should have been independently
tested with water at high temperature to verify it was compatible before it was used for testing.
2. It would have been better to keep the test configuration as simple as possible and not test multiple technologies
concurrently. The spin thru centralizers should have been tested separately from the pump.
3. The drive system should have been larger (i.e. oversized) to ensure that the system was never power limited by the
VFD or the drivehead, especially when there was uncertainty in the peak torques required for start-up. Better to have more horsepower, than not enough.
4. The upgrade to the loop should have included making the casing more resistive to radial forces, as it was known at
the time that large volume lift systems were to be tested, with the potential for vibration.
Testing of Metallic Stator PCP – Round Two
The results of the first round of testing were analyzed, which included constructing a vibrational model of the casing section (i.e. using Finite Element Analysis) to identify ways that the loop could be modified to account for the possibility of system vibration, while still allowing for thermal expansion. A go forward plan was then formulated, which involved not only modifications to the loop itself but to the system completion as well.
Modification to the Installation and Experimental Flow Loop – Round Two
The key changes that were made to the experimental setup, downhole and surface completions are summarized below: