Your comments are right on target with respect to load saturation curves. The industry has a lot of training to do. From my experience, not only do MOST operators not know enough to ask, but those who do are in the offices and not getting the information to those making settings and running the pumps.  Talk about production optimization! Where is everybody on this topic who are on the optimization bandwagon?

About oscillations - an idea of a scenario is provided below in dark blue  - However, no snipes unless we get data shared on this board. If you are able to provide various data such as well data, also saturation test, heat rise and lock rotor test for different motor sizes on this board please let us know, - then we can use this as a basis, and go from there. Will need good data from a manufacturer or operator - just "anecdotal" evidence is not good enough and the snipers have a field day discrediting. Also, can previous data be provided from slurry loop tests?

If all this type data is proprietary, then regrets unless the operator or owner allows.

Quick analogy - when I joined the esp industry, I worked for one of the larger manufacturers.  I started in what was called the "horizontal" pump group, which is nothing more than an esp that is mounted on a skid coupled to a surface motor or diesel/gas driven engine, and a bearing housing, with a welded intake upstream of the bearing housing and a flanged discharge to the pump.  The industry and that particular company had been faced for years with short runs and what they referred to as "thrust chambers"  that had continual problems. It was so bad that people were calling the pump a horriblezontal. Everyone seemed to fret about the "thrust chamber".(in the early days they just mounted the pump on a pedestal and without a bearing system - I suppose they thought that the motor bearing could handle all that pump downthrust !)  Over the years, they had concocted all sorts of patches to try and remedy the problems - blowers and heat exchangers to keep the "thrust chamber" cooled, studies on shaft alignment and vibration, mounting procedures - all sorts of guesses to stop the problem.

All modesty aside, I asked in the first week or two - "what is the L10 life of the bearings?" And I got blank stares. Anybody with ordinary skill in the pump world knows about L10 bearing design. But the ESP world is different, and you would not necessarily expect someone with all their career invested in downhole systems to understand bearing design. In fact, a bearing in the downhole esp world means something different than to engineers and operators on surface, but that's another chapter.

Getting back, everyone seemed to be an expert on the thrust issues and ideas to solve it, but not a single person knew anything about L10 life and its correlation to proper bearing design!

A quick call to Timken and we worked out the L10 life one afternoon for this pump company's "thrust chambers".

Guess what? It was about 28 days! Can you believe that? This pump company had worked for many many years on these bearing system problems,and the best solution they had actually carried a L10 bearing life of only 28 days!. It's no wonder they had issues!    So that was ground zero from there on. Do you think that the pump company wanted to tell everybody once they learned their design life was only 28 days ? 

And anytime there was somebody with a great bearing solution idea, it always went back to L10 first. If I were buying one of those pumps, I would first ask to see the L10 calculations for their bearing system - unless of course they were utilizing a non ball bearing  like Glacier brand bearing system. Regardless, bearing design life - whether L10 or other should be required. I would NOT trust a written value unless I saw the calculation! I bet you'd still get blank stares from most of the salesmen if you asked them the L10 . And guess what? L10 bearing life is still not even published information for any of their horizontal pump systems utilizing roller bearings to handle down thrust. As important as downthrust is to the reliability of a surface esp pump system, L10 bearing life is not published. Why not?

Point is that you need real data. Without real data, what's the use?

Back to one of the original questions, it was " Could you please explain the relationship of sand bridging to shaft oscillations.  It is not intuitively obvious to me.  Some anecdotal evidence might be useful." How about this scenario? :


"Scenario"
Pump starts with high fluid level, high flow and loaded enough to run smoothly.

As the well pulls down, the intake pressure and flow drop along with the load on the motor.

Some motors start oscillating. Note that some use the excess energy thermally and begin to heat up the windings, cable, connections, inverters etc. depending on the motor design.

When shafts oscillate, this causes loss of flow, probably due to loss of RPM's but some may vibrate so much that the operator actually loses pressure and flow across stages.

During oscillation and low flow, the well recovers until you would have enough intake pressure to stabilze the shafts and then very rapid pull down of fluid level occurs-(may only be 10 to 25 PSI cycle but constantly changing- important note: those with lots of experience will tell you that many chart for weeks at a time and never see this oscillation on their charts, so should chart PSI over three to four hours to really see what is happening).

When the level is low enough and load is light enough, cycle starts all over again.

This constant change in intake pressure (formation pressure) prevents sand bridging that you would normally see with a slow steady draw down.

This cycling also causes constant load changes on any device monitoring current changes to try and control voltage output (generators and VSC control boards).

This also makes it extremely difficult for pressure sensors surface packages that need to send a constant DC volts downhole to interpret what it needs to send or what readings it is receiving.

Indeed, those with this type of experience will tell you that many sensor problems over the years weren't really sensor problems at all.

Comments & well and equipment support data would be appreciated in line with this scenario. Challenge to the pump vendors to step up to the plate.

David,

Could you please explain the relationship of sand bridging to shaft oscillations.  It is not intuitively obvious to me.  Some anecdotal evidence might be useful.

Regarding voltage and load on motors.  To truly determine what voltage should be applied an operator could request  load saturation curves for the motor.  This is discussed in another conversation on this board. Some manufacturers nameplate data do not reflect the optimum voltage for a fully loaded motor due to other considerations.  Load sat curves show the motors perfromance for a variety of load and voltage conditions. 

If these are not forthcoming an operator could simply "play" with the voltage and watch the current draw.  Adjust the voltage upward or downward.  A drop in amperage indicates you are adjusting in the proper direction.  Continue until the current starts to rise again.  This will give you the optimum voltage for the load condition the motor is under.

Regarding sand bridging, there's plenty of evidence with respect to oscillations. However,  I am not with Centrilift, Reda,  Wood Group or other manufacturer, so I cannot share the data which belongs to them. I know it exists though.

As for why motors often end up with too much voltage,  here are a few comments to think about:
 
1. Voltage required is the voltage required for the horsepower load. Do you have access to Autograph or other sizing program? The only true way to know this is to go to the H-Q  tornado curve and see what the HP/stage is and multiply the number of stages by this number. Then add any extra HP for energy used by seal and/ or gas separator. In some cases of heavy fluids being produced you may need to add HP for fluid density but usually we are producing fluid lighter than water so this is seldom a concern. Formula-Req. HP/Nameplate HP/square root*NP volts, gets you close. Just adding volts for voltage drop to get surface volts will never be correct unless HP Req. and NP HP match exactly.
2. A drive changes voltage by a ratio. Going from 60 to 50 HZ the voltage change would be 50/60 (.833) * 60 HZ volts. The HP change from 60 HZ to 50 HZ is to the cube and this is a big change in HP required. The ratio change never drops the voltage quick enough. Because we are lowering the load and the volts many don't think you can have too much volts to the motor. They are wrong, the motor only requires the voltage needed for the load. The drive ratio change is only close enough if the HZ change is 2 to 3 HZ. The ESP drive guys don't want anyone to know this.
3. Most all pump applications and sales people always start with nameplate volts and add voltage drop. Have you ever seen a motor selected that wasn't more HP than what is required? If we start with a motor that has more voltage than required then add voltage does this make sense? If the motor is larger than required the amperage should be less than the nameplate amps. Only motors that are close to the required HP  will show nameplate amps and most of them actually have less volts than reqired due to not adding impedance of the transformer. Quite often people say the motor is 100% loaded because it is running nameplate amps. I can almost always get another 75 to 100 BPD out of these. We'd be pleased to provide training to operators and  teach classes with some experienced field service professionals on operating within the motor characteristic curves. Question: how much training do your operating personnel receive in this regard?

Are you an oil & gas operator? How much extra would that 100 B/d be worth to you?Multiply that 100 b/d x the price of oil. Even at today's levels of $ 35 b/d, that translates to $3500 day for a typical well. On an annual basis, that's about $1.3 million dollars PER WELL. In addition, the two of us can show your operations and engineers the nuances and types of questions to ask esp vendors to get your production up and maintain reliability.

Theoretically I can see this happening (resonance) but in practice how often does this come up as a cause of failure?  Any data to support this?

Also I assume you disagree with him then on lightly loaded shafting systems causing oscillations?

I know the guy whose father-in-law owns Green Country.

Interesting exchange. I would like to add a basic thought . And that is to make sure it is agreed first as to what the problem(s) really are, before trying to understand and solve them.

Looks like one school of thought is to prioritize understanding of kVA in the context of selecting where on the system curve to run the pump.

The other school of thought has the perspective of actual cause and affect of amperage (loading) re: oscillations.

Another comment seems to be about motors being oversized to begin with as being the problem, which sounds to me like the problem identified is the applications process and perhaps those engineers not having enough time invested to understand real experience downhole operating conditions, versus their tools to solve problems (ie the pump system models and calculations embedded in their sizing software ( like Autograph).

Tell me how I can help in this dialogue. I would like to learn and get involved too.

I agree, and I would like to add that it's important to make sure in advance to let the lab know what you are looking for with the sand analysis. - Measurements of course  AND ALSO photos for criteria like sand angularity etc , - make sure to get sharpness and grain size measurements, compound compositions %, etc.

Gentlemen

Unfortuneately (at least in the short term for those affected), layoffs and restructurings are
are consequence of the business cycle and decisions to keep companies well capitalized and profitable by their managements. Let's remember Business 101 -  at the end of the day it's all about ROI.

As long as the relatively well-capitalized companies handle it professionally and fairly, let's cut the companies some slack. The ESP manufacturing industry I believe does their best for their employees in the face of lots of fast paced changes.  --- Like those seen now.

Those are my viewpoints. I also appreciate hearing about changes noted above. Healthy exchange and keeps us updated as to what's really happening in the industry re: personnel etc.

Thanks.

Further evidence of this is ALL job openings on BHI website in Claremore/Tulsa for C'lift were pulled down.

wrong about CLIFT. Were OA+, but mothership Houston switched to EV. Listen as mgrs are now putting together new budgets and head count cuts for early 2009 - mandatory. Here we go again, kids to learn reality? merry xmas.