Skid Mounted Pumps 2

[bones206]

I am a (1st year) structural engineer checking the design of an industrial skid, which includes two 4x6-15 centrifugal pumps for circulating hot water and two large heat exchangers. The pump and motor are mounted to a single bedplate, which is bolted to the skid frame. My concern is that in checking the design calculation the engineer did not consider vibration of the pumps. Since I have no experience with pumps I was hoping to gain some insight.. I have asked mechanical engineers in the office and they say that generally they prefer a pump to be mounted on a stiff massive base - usually concrete. They said that steel frames are less desirable because of deflections of the pump supports and long term fatigue from the cyclic loading. However, they could not give me any solid guidelines for designing a steel skid, just that it is 'not preferred'. I have read through a few handbooks on pumps but when they refer to foundations, they assume a concrete or grouted foundation. They give rules of thumb for the ratio of the foundation mass to the pump mass but I believe those rules of thumb are based on stiff concrete pads, and would not apply to a flexible steel frame.

What should I check for in terms of stiffness, vibration and fatigue? Is a vibration analysis required? Are there locations where stiffeners are usually desired?

Should I be concerned about thermal growth of the skid causing misalignment of the pump components, or will the bedplate keep everything in alignment? The temperature ranges from below freezing (during shipment) to around 150 F during operation.

Thanks in advance

 

[BigInch]

It usually depends on the size of the pump/motor as what you have to do. Many large pumps and compressors all come mounted on skids from the factory. Skid interior framing beams should be directly under pump bearing plates that are attaching to the skid so the pump can be indirectly supported by the foundation below through the framing beams. Those framing beams should be grouted to the foundation rough concrete. If the pump is large and vibrations are critical, you should use epoxy grout, however smaller units could use stiff masonry grout. Smaller units may have interior beams that do not contact concrete below, if they are light enough and the frame can transmit all forces to the skid rail framing members.

What's the HP rating, the weight of the pumps, weight of motors and pump rpm? If less than 150 HP, you should be able to get away with a lighter framing to skid rails. Over 250 HP, you would usually want to have grout contact with the interior framing members. Over 500 HP, you should definitely use epoxy grout.

The COE has a guide to design foundations for large vibrating machines. I think that it has some vibration limit advice too. I have attached a copy of it here for you (its a self-extracting exe file that turns into a PDF).

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[bones206]

BigInch - Thanks for the guide and the advice

It's a 60 HP rated unit and operates @ 1475 rpm. The total weight of the pump and driver on the bedplate is around 700#. So it sounds like a small pump and I shouldn't worry too much.

The skid frame is about 20' by 12' by 12'... a big box. The upper level is needed to support pipes because the concept is that the entire skid arrives at the site as a totally pre-assembled plug and play module. The supports for the module are 6 columns that are welded to embedment plates pre-installed in the plant. The skid platform is elevated so that the bottom of steel is about 6" off the concrete slab. The module will likely be located at an elevated slab with other mechanical equipment (on similar skids).

Since the skid will be elevated 6", I'm not sure if it would be feasible to grout them. Also, the engineer who designed the skid said that he doesn't want to grout.

The vibration still worries me. This pump and module are supposed to last 60 years. Is fatigue an issue? What kind of cyclic loads are transmitted from the bedplate anchors to the steel? The pipe supports are rigidly attached to the module frame and both pumps share the same framing for pipe supports.

The skid framing directly below each pump is basically 2 parallel WF beams with 3 angles as cross members, located under the ends of the bedplate and under the c.l. of the coupler. It seems like it would be sufficiently stiff but not very massive. I also wonder about resonance with the pipes and frame.

 

[BigInch]

If the engineer that designed the skid does't want to grout, then he should have made the skid stiff enough to carry the loads to the support points without undue vibration. If that's OK, well ... then its OK, no grout.

Fatigue could be an issue for the pump. 60 yrs is a long long time for a pump, but if the beams and frame are lightly loaded, probably not too significant.

As I understand this, there are two ways to minimize vibration, add mass and/or stiffness. Mass is more conveniently added in the form of a large concrete foundation and reduces amplitude of vibrations, since more force is needed for the same travel of distance in amplitude. Stiffness reduces vibration frequency, since forcing functions are kept farther away from resonant frequencies and amplitudes are thereby reduced. Since too much mass above grade is difficult, perhaps you are limited to adding stiffness. If the pumps made large forces, keep the resonant frequencies of the beams and pipes away by at least 30% either side of 1475/60 Hz to get reasonable performance, but being small and centrifugal, they probably are not very great at all. Recips can be another story. I think if you designed the beams for motor and pump torques due to motor short circuit loads, that would be more than enough for running loads too.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[bones206]

Thanks again BigInch. That was exactly the type of guidance I've been looking for. Much appreciated.

 

[pennpiper]

I support what "BigInch" has told you.
I would also like to add a little that might ease your concerns.
There are lots of pumps that are installed in or on structural frames (not on concrete foundations) that do just fine.

Think about ships all sizes. Aircraft Carriers with Bilge pumps, Fuel Transfer/Loading/Unloading Pumps, Boiler Feed Water Pumps, Desalting Water Pumps, etc.

Think about Sky Scrapers (tall office Buildings) with HVAC system Cooling Water circulating pumps located on the roof.

Think about off shore platforms. They have pumps located on the steel frame platforms.

It can be done, because it is being done.

 

[Artisi]

My suggestion would be to get on your bike and visit a number of pump installations of similar application or visit a few pump companies that supply and mount pumps to get some first hand knowledge, at least then you will know what you are reviewing.

 

[BigInch]

Artisi, I've found mfgrs arn't too cooperative in giving foundation and support framing advice. Apparently they think that they will be held responsible for any future malfunctions that can be blamed on vibrations of any kind, so I'm sure you can see their motive.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[Artisi]

BigInch, I was suggesting to see "how it is done" not necessarily to ask for concrete detail (no pun intended)from any third party - which seemsto be what the OP is looking for now

 

[TenPenny]

A more-or-less unrelated point, but one that came to mind: put a bit of thought into maintenance - with a 60 yr planned life, I hope the skid designer put some thought into how the pump and motor can be removed for maintenance / replacement. Since you mentioned 'a box' and piping overhead.

 

[desertfox]

Hi bones206

I have done a lot of pump motor skid design mainly for waste water industry.
The pumps we used were always rotatory and the impellers were always balanced and the units were tested before dispatch so vibration wasn't a main concern.
Most of the skids were however grouted in and to my knowledge the skid base was fully supported although I can remember that some skids were mounted on frames but even then the longnitudinal members and the two end cross members were supported.
The method we used to size and design our frames were as follows:-

1/ Calculate mass of pump and motor.

2/ Select longnitudinal member section
in our case we used 'C' section.

3/ Stress member as a simple supported beam using
half the mass calculated at 1/ and check
deflection. Our safety factors were about 3 or 4
so stresses and deflection were quite low.

4/ We then drew the frame up based on this an the
width would be governed by the physical sizes of
pump and motor.

We use to calculate the size and number of bolts by considering the maximum torque the motor could produce and applying that across the width of the frame, ie looking on the end of the pumpset and imagining the frame twisting up about one side. The bolts were then usually placed along the longnitudinal section of the beam with equal pitches or as equal as possible using the centre line of the pump impeller casing as a datumn.
albeit that this method seems quite crude all I can say is that it worked and the company actually started producing pumps around the late 1800's

 

[bones206]

Yes, the designer of the module designed the overhead framing for hoists or chainfalls to perform maintenance on the pumps.

This module is for a nuclear plant and according to the experienced engineers in my office, equipment is usually "stick built" on site (on concrete pads) rather than fabricated on a module off-site. This particular designer does not have experience designing a steel support frame for vibrating equipment.

I actually did consult with the pump manufacturer and they were less than forthcoming. The one point they stressed was that the bedplate of this particular pump was specifically designed for a grouted foundation. They did not want to discuss the implications of the pump being mounted on a skid other than saying it should be a "sound, engineered design".

Artisi - great advice. I'll see if there are any local pump installations or manufacturers that I can check out. If there are any skid fabricators on this forum I'd love to hear from them!

 

[Artisi]

desertfox - the late 1800's company wouldn't have had the initials AC would it?

 

[desertfox]

Hi Artisi

Not sure to be truthful its a while since I was there, I don't think they did.

desertfox