improving progressing cavity pump … been used, from monitoring torque to manual fluid levels. to...

kudupump.com

IMPROVING PROGRESSING CAVITY PUMP PERFORMANCE THROUGH AUTOMATION AND SURVEILLANCEK. A. Woolsey, SPE, KUDU Industries Inc. © Copyright 2010, Society of Petroleum Engineers

KUDU Industries Inc. Pg 2

Abstract To extend the run life of the pump while producing all available fluid is the goal of all PCP operators. The primary challenge is to do so without starving the pump and causing damage to the stator. The petroleum industry has been searching for years to find a reliable way to control progressing cavity pumps for pump-off. Several methods have been used, from monitoring torque to manual fluid levels. To date, none of these have been commercially successful. A method for controlling these wells has been developed combining wedge meter flow technology and microprocessor control of both electric motors using variable frequency drives and hydraulic motors using proportional control valves. This method has proven to

be accurate and reliable, extending the run life while producing all available fluids.

Combining this automated technology at the well with a web based system that feeds back real time data to a dedicated SCADA host allows PCP technical experts to diagnose problems, and operators to respond quickly to changing well conditions.

This presentation discusses the advances in automation and optimization of progressing cavity pumps. The method will be explained and data will be presented of field study results showing actual well tests.

IntroductionThe acceptance of progressing cavity pumps in the oil and gas industry has grown throughout the world. One of the major concerns has always been causing stator damage if the well is pumped off. The most common prevention is to ensure that there is a substantial amount of fluid level above the pump. Operators taking manual fluid level shots, and adjusting pump speed has been the most widely used method to maintain a safe fluid level. With advancements in fluid flow measurement and proven algorithms, the PCP

controller has eliminated this concern, while producing all the available fluid from the well bore without damaging the pump.

With this automated technology and a web based SCADA system we were able to monitor several data trends and discover additional control algorithms. Continued development of these algorithms is ongoing.

Pump off Control for Progressing Cavity PumpsTwo main objectives were defined when the theory of the PCP Controller was first introduced. The first objective was to reduce premature pump failure. The most common failure is stator damage due to pump off. Pump off is defined as a lack of fluid entry into the pump. This causes a lack of lubrication to the stator resulting in extremely high temperatures being generated. The high temperature ultimately burns the elastomer in the pump. The rubber surface of the stator becomes hard, brittle, and cracked. In severe circumstances, the stator contour is completely torn up producing rubber at surface. This may be caused by one, or a combination of the following, problems; plugged pump intake, poor inflow, or production rates exceeding inflow.

The second objective was to automatically optimize the well. There are several problems associated with trying to maintain a safe fluid level with the use of manual fluid level equipment. Fluid levels are only an indication of how much fluid is over the pump at the time the fluid level is taken. There are a large number of misinterpreted fluid levels due to surface and rod noise as well as foam. Downhole anomalies such as high liner tops, bad dog legs, or other “non standard” configurations will result in false fluid levels. It is extremely time consuming to shoot fluid levels, adjust speed, allow the well to stabilize and shoot the fluid level again. This manual process is often not a quick enough reaction to account for changes in inflow.


By optimizing the well, we not only eliminate the fluid level monitoring problems above, we also increase overall production. This is accomplished in several ways. First, by keeping the fluid at or near the pump

we are maintaining the lowest bottom hole pressure promoting the most fluid inflow. Second, by being able to pump over capacity the system is able to make a quick recovery from downtime.

Development of Theory and Control AlgorithmsThe primary control algorithm of the PCP Controller is the production control algorithm. Its purpose is to optimize production of the well without causing pump off. The patented control algorithm varies the speed of the pump in steps while measuring the amount of fluid production from the pump. As long as sufficient fluid is available at the pump intake to completely fill the pump cavities, a linear relationship occurs between fluid production and pump speed. The special hunting algorithm continuously looks for the most production. The controller continues increasing the pump speed in user-defined steps, measures production rate at each step after the settling time, and establishes a speed/rate relationship.

A description of the production control algorithm is as follows:• The PCP Controller begins an operating cycle at a user-

defined startup speed. When the well is first started at the suggested start up speed, for example 150 rpm, the initial well state is the settling period. This is the period of time that the production from the well is allowed to stabilize after any speed change. For our purpose we will use three minutes for the settling time.

• After the three minute settling time the controller goes into sampling period. This is the period of time that the production from the well is measured and a flow rate is established. For our purpose we will also use three minutes for the sampling time. The flow rate that is established is total fluid in barrels per day for that period. The pump is then sped up between the hard set minimum speed change and the maximum allowable speed increase that is preset. The well is then placed back in the settling period again.

• After the three minute settling period and the three

minute sampling period, the algorithm looks at the flow rate and determines if it satisfies the minimum production increase requirement that was preset at commissioning. For this purpose we will use 5%. If the production increase is greater than 5% the speed of the pump will be increased a percentage of the previous production increase while remaining within the minimum and maximum speed change window and the hard set minimum speed change. This process is repeated until the minimum production requirement is not met, or the pump reaches the maximum speed allowed. At that time the pump speed is reduced to a value between the maximum and minimum speed decrease window that is preset at commissioning. See Figure 1.

• After the well is ramped down in speed the settling and sampling periods are again observed and the production still needs to meet the minimum production requirement. There is a number of consecutive speed ramps allowed that are configured at commissioning before the challenging algorithm is implemented. For this purpose we will use three. After the well is ramped down in speed three consecutive times and the minimum production requirement is not met, the pump is sped up the hard set minimum speed change, and the settling and sampling periods are again observed. If the minimum production requirement is met then the pump continues to be sped up and the production is tested. If the minimum production requirement is not met then the pump will be ramped down again and the process continues.

The PCP Controller also has secondary control algorithms that monitor the following:• A polished rod torque value either from the electric

variable speed drive analog output, or hydraulic


pressure from the hydraulic pump to the hydraulic motor. This value is used to protect the rod string and pumping equipment. Two violation settings are provided for torque with the option of programming the well to either shut down or slow down when the limit is exceeded. The torque limiting violation is used to limit the rod torque. The current torque is compared to the specified limit value to detect if the pump is beginning to show evidence of stalling or fluid starvation that can cause friction between the stator and elastometer, thus increasing the torque. Torque is monitored every second, and once it is violated, the PCP Controller will decrease the speed by the specified torque limiting speed value, and continue reducing speed until the minimum operating speed is reached. In this state, the PCP Controller either operates in a downtime or violation minimum running speed mode. The Critical Torque violation is used to either shut the well down or run in minimum speed when in a high torque situation. The current torque is compared to the specified limit value. This is an instantaneous action that reduces the speed to the minimum operating speed or shuts the well off depending on the action selected.

• A current flow rate value either via the wedge meter or another trusted flow measurement device. Two violation settings are provided for flow rate with the option of programming the well to either shut down, or slow down when the limits are exceeded. A low production fluid rate limit is specified defining the minimum amount of fluid production allowed. Production for the current sampling period is compared to this value to detect a fluid starvation/dry pump condition. This feature also helps to detect (at the rate of one-minute resolution) if something is wrong with the production transmitter. When a violation occurs, the PCP Controller decreases motor speed by the value specified. At this state, the PCP Controller either operates in a downtime or minimum running speed mode.

• A low pump efficiency limit is specified defining the minimum efficiency the pump is allowed to operate at. Production for the current sampling period is compared to this value to detect low pump efficiency. When a

violation occurs, the PCP Controller decreases motor speed by the value specified. At this state, the PCP Controller operates in either a downtime or minimum running speed mode.

All secondary control features slow the pump when one or more set point limit is violated. If the monitored variable continues to be in an alarm condition when a programmed minimum speed is reached, the PCP Controller will stop the well to prevent damage to the pumping equipment.

After proving the initial production control algorithm it became evident that the flow meter was a critical part of the complete system. Although exact accuracy was not necessary, repeatability was critical. Many types of meters were tested, however, the wedge meter proved to be the most reliable when measuring flow at the wellhead. Working in conjunction with the wedge meter manufacturer many improvements were made and this device is now the meter of choice for most applications. See figure 2.

Combining this automated technology at the well with a web based system that feeds back real time data to a dedicated SCADA host allows PCP technical experts to diagnose problems, and operators to respond quickly to changing well conditions. See figures 3 & 4. Utilizing statistical process control the SCADA host allows operators to determine when changes in field conditions occur. By monitoring the difference between the maximum allowable working speed of the pump and the current speed we can immediately determine if the well is optimized. Automatic reaction to changing field conditions prevents pump damage. See figure 5. This has shown significant results in increased production, and pump run life, as well as a reduction in maintenance and down time, thus providing better cost/time management for operations.

Continued improvement with customer input has driven development of additional algorithms that make producing wells with a high sand content much easier. See figure


6. With additional equipment at surface, this technology has successfully performed automated flushes of large

amounts of sand without operator intervention. See figure 7.

Field Study Results and Historical Data1. A 15 well pilot project started in January 2007 included

the host SCADA system and the PCP VFD Combo unit complete with wedge flow meters. The results showed an increased average oil production from 120 m3/day, to 145 m3/day. An unexpected benefit of increased average gas production from 19 to 23 dec/day was also noticed. This showed an estimated payout in 120 days. See sample well from this project on graphs 1 & 2.

2. A single well pilot started July 2008 was chosen due to its extreme production difficulties For the first six months prior to the installation of the PCP controller, this Bakken zone well had four pump changes due to sand and inflow issues. The well produced a total of 2711 m3 with an operational cost of $198,000.00. On July 1st, 2008 Kudu installed a 300TP1200 pump and added the hydraulic PCP controller with wedge flow meter. For the next six months the well produced 19,779 m3 with no work overs. The pump ran for 14 months before it needed to be changed out due to wear and tear from abrasion. The hydraulic control system expertly maintained speed changes using PID loop

control on the proportional valve without adding heat to the system. See graph 3.

3. A single well pilot started in Dec 2009 was chosen by the producer because they were unable to keep the well running. Before optimizing the well the concentration was on producing the sand that was preventing the well from running. Once this was achieved we were able to increase fluid production. By bringing the well on slowly, limiting the maximum speed over four stages the sand was cleaned up to a very manageable level and the revenue for the well was increased from $ 0/day to $7,798/day with an oil gain of 19.4 m3/day. See table 1.

4. A single well pilot started in Sept 2009 was chosen due to the short run life of the pumps. Prior to installing the PCP controller the pumps were being replaced every 3 to 4 months. A new pump was installed at the same time the PCP controller was installed. This pump has been producing since Sept 2009 and the most recent test showed a pump efficiency of 74% with an oil gain of 1.84 m3/day, increasing oil revenue by an average of $925.00/day. See graph 4.


GRAPH 1: Example 1 Results




TABLE 1 Example 3 Results

Nov 26 - Dec 26/09

Pre-controller #1

744

0.00

0.000

48.97

0.00

$ 0.00

$ 0.00

Hrs On

m3 Oil/Day

m3 Sand/Day

m3 H2O/Day

m3 Oil/Day

$/Day

$/Month

741

0.99

0.006

38.45

0.99

$ 398.11

$ 12,341.40

744

1.29

0.000

108.95

1.29

$ 517.41

$ 16,039.80

744

10.19

0.000

139.40

10.19

$ 4,097.81

$ 127,032.00

744

19.40

0.000

144.30

19.40

$ 7,798.80

$ 241,762.80

Dec 27 - Mar 15/10

Post-controller #1

Mar 16 - Apr 15/10

Post-controller #2

Apr 15 - May 15/10

Post-controller #3

May 16 - June 15/10

Post controller #4



ConclusionThe PCP controller has proven to be a successful and viable well manager for progressing cavity pumps, in meeting the objectives originally set forth. Overall productivity has been increased for the well and for the entire field by increasing daily fluid production, recovering more quickly from downtime, and enhancing cost and time management for operations. The PCP controller meets

and exceeds the goals and expectations of maximized production, minimized downtime, reduced equipment failures, and extended equipment run life. By eliminating the need to continuously monitor fluid levels and manually adjust pump speed, operators are able to concentrate on other tasks. Secondary control features allow protection from over pressure, high torque, and sand plugging.

AcknowledgementsI would like to express my gratitude to Lufkin Automation and Theta Oilfield Services Inc. for their product

information and equipment used in this paper, as well as IHS for the Accumap production and well data provided.


FIGURE 1: Minute by Minute Production, Speed, and Torque Trends

FIGURE 2: Wedge

FIGURE 3: XSPOC Group Status Screen


FIGURE 4: Well Status Screen

FIGURE 5: Minute by minute production and speed trend showing automatic reaction to increased flow line pressure.


FIGURE 6: 7 Day trend data showing automated sand handling

FIGURE 7: Trend data showing automated flush technology

+1 403 279 5838kudupump.com

improving progressing cavity pump … been used, from monitoring torque to manual fluid levels. to...

Documents