Electric Motors driving Sucker Rod Pumps (oil pump jacks): correctly measure their Power consumption

[mrspcs]

We need to measure the Electric Power consumption (kW power intake) of electric motors driving Pump Jacks in oil wells.
I am pretty sure this requires the use of a special type of meter due to the cyclic nature of the pump operation (along its cycle it goes from high to low % Load values). Something like the Dranetz Demand Analyzer (e.g. take readings over a certain time interval).
Common hand-held, single digital output devices (e.g volts, amps, kw etc) won't do for this particular application.
I have seen this issue documented in a few papers (e.g. from the Society of Petroleum Engineers -SPE-).
I would appreciate if someone can indicate specific references regarding applicable standards for this measurement (like API, IEEE, etc).
Comments and suggestions on this issue will also be greatly appreciated.

 

[DRWeig]

Here are a couple of papers from IEEE Transactions. They aren't standards, but the first one seems very useful for you.

There may be standards for measurement of power in protective relaying systems -- those would be much faster than you need as well, I would think.

Using power measurements to diagnose degradations in motor drivepower systems: a case study of oilfield pump jacks

Microprocessor based fast active power measurement

If it were me, I would just get the Dranetz machine, or one of the LabView apps. They are expensive...

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[davidbeach]

A decent motor protection relay ought to be able to provide what you're looking for.

 

[itsmoked]

mrspcs; Thanks for the clarification.

If you want the most accurate measurements for the power put into the well verse the oil volume returned then I repeat:

The most accurate method is not fancy instantaneous instruments it is merely a standard WattHour ENERGY meter. Energy divided by time is POWER. The standard meter gives the REAL POWER. That's the power that you pay for. It's the power that lifts the oil. The more time the energy is measured over, the more accurate the measurement, the more accurate the resulting power measurement. Standard watthour meters are fully capable of tracking the power excursions of a sucker pump motor. They border on the trivial, as far as watthour meters are concerned.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

It's all about the money. Learn to take spot readings on the meter that is responsible for the billing.
I suggest an exercise to get a feeling for motor efficiency on pump jack service.
Take the published efficiency curves of several suitable motors. As well as different efficiency rated motors, use the curves from motors that are one or two sizes over-sized.
Estimate the work curve of the pump jack cycle.
Set up a spread sheet to calculate the efficiencies of the motors throughout the pumping cycle and then calculate the overall efficiency of each motor.
It may be interesting to consider the cases where the counter weights are more than or less than the optimum mass.
Over the years I have found this type of exercise valuable to develop a real "feel" for an effect, and an aid in developing personal "Rules of Thumb". I seldom do it more than once. The first time I did power factor correction, the calculations took several hours. The second time the calculations took several minutes. But I have never regretted the time spent on the first calculations.
This exercise will probably take several hours, but at the end of the day you may have a better feel for pump jack motor efficiencies than most people.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[electricpete]

HOWEVER, directly measuring the Power intake or the variables needed to calculate it (V, I and PF) in the case of a highly variable load like the Pump Jack.....
In loads of this type, the I and PF are varying constantly and rapidly. Hence the need for this particular type of instrument.

or any other aspect of this task

I have no standards. Don't beat me up for responding.. I wanted to point out a potentially very tricky aspect when using digital instrument in the presence of highly varying loads (potentially aggravated by resonant torsional swings). If there is a period when the oscillation causes the real power to reverse, then during that period it contributes to increasing the integral of apparent power while decreasing the integral of real power. Thus any kind of "power factor" calculated as a time-averaged real power over a time-averaged apparent power would be meaningless (at least much different than anyone would expect). Example was discussed here:
thread237-249262



=====================================
(2B)+(2B)' ?

 

[LionelHutz]

How I inteprete what you're writing makes no sense. "Tell me what kind of meter to use to measure the power intake to give a fixed number so I can calculate the efficiency of the motor."

The efficiency of the motor varies as the load changes. Coming up with some type of time averaged load and efficiency to compare 2 motors won't give you any true answers.

There have been multiple posts suggesting to use kWh and divide by time. Those answers were all 100% exactly correct since that calculation is the number you've now posted that you wanted 3 times. It's turned into listening to a skipping record here...

Should anybody question any of that, I'd suggest taking the case up with the authors and/or organizations providing the material. I am merely a reader/user of it.

I don't think the authors of the papers you're reading would appreciate being blamed for your lack of understanding.


On another note, those steps to calculate the percentage load on the motor are good for nothing more than a rough estimate. It completely skips the part where the motor efficiency has changed once the load changed. Then that paper also seems to imply that you first calculate the percent load on the motor using the motor rated efficiency and then use that answer to calculate the efficiency of the motor. I'll save you the bother of doing any calculations. The efficiency answer will be the motor full load efficiency. You'll need a different method if you want a useful answer.

If you want to figure out the efficiency of your pump jack then use gallons or litres of oil pumped per kWh of energy used.

 

[cranky108]

If the goal is to optomize the motor size, be aware that as the load reaches the motor rating the motor speed will slow down, and there is a stall limit on how much you can put on the motor during that peak. In other words, if the pump is balanced you can optomize better than if the pump is unbalanced. For an unbalanced pump you will be required to oversize the motor for the peak loading.

Some of the problems you will run into with pump jacks, is some operators use old or rewound motors, and even if they are in good shape, the nameplates may be unreadable because of age or paint or both. So even if you think you know the size of the motor, it may not be so.

 

[DRWeig]

Just curious here -- I am a pretty good power meter guy on the lower-tech end, my company develops and markets to the building automation systems world. Nothing fancy there, more of a workhorse with good enough accuracy to match the utility's meter on a building. High enough sampling rate to meet the accuracy class, something over 4000 samples/second. Lots of data logging memory.

However, the abstract to the paper I cited in my last post indicates that something more is needed for the OP's purpose perhaps:

A study of diagnostics of faults in motor drive-power systems is reported. The study focuses on oil-field pump jacks and includes basic analysis, computer simulation, full-scale laboratory emulation and measurements in the oil fields. [highlight #FCE94F]Spectrum analysis of instantaneous power and current delivered to the electric motor is used for diagnosis[/highlight]. An original electronic system for conditioning the input instantaneous power and current is explained. A model for the pump jack and its common degradations is given. A full-scale load control system emulates measured pump-jack loads in the laboratory

I didn't want to plunk down $13 for the whole paper -- it's the OP's pump jack issue, not mine. Note that I also recommended measuring with a simple device earlier in this thread. The energy portion would be fine; therefore, the average power will be fine as well. "Spectrum analysis of instantaneous power" is certainly not in my game. Thoughts?

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[LionelHutz]

Dave - it's used for fault detection. The OP made no mention of wanting to do this.

 

[DRWeig]

Thank you Lionel. I am bad about re-reading the OP when a thread gets long...

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies