Support for Solar Boiler

[sullypape]

What is the maximum length of supporting steel structure for support of an 1800 foot long solar steam boiler? Any special considerations in addition to considering temperature delta for air and process?

 

[JStephen]

Assuming you're talking about something like this:

http://www.rechargenews.com/energy/solar/article194840.ece

I don't really see that there would be a limit as to the height, it would just be an issue of working out the most economical and practical configuration. You could always make it higher if someone was willing to pay for it. For example, the Margaret Hunt Hill Bridge going up in Dallas has a 400' high center arch.

The geometery is not going to be something addressed by normal building codes, so wind loading will not be standardized. Also, you'd have potential vibration effects in the wind that you'd need to consider.

 

[sullypape]

Picture pretty much sums up the structure. Concern is that the piping and overhead structural supports extend 1800 feet without any expansion joints. Practice and available literature for pipe racks has always indicated a maximum structure length of 400 feet between expansion joints. With the temperature delta and structure length, and piping and structure anchored at midspan by braced frame, the end frames would be pushed out of plumb!

Height of supports constant for each pipe run 60 to 120 feet high.

Any thoughts?.

 

[JStephen]

That clarifies the question considerably. Unfortunately, I don't have much more to offer, either. It seems you could go ahead and put in expansion joints, you could go ahead and design the structure where it could be pushed out of plumb, for that matter.

Current practice on railroads is continuous welded rail without expansion joints. But there, they have more lateral support available. And I think they try to lay the rail so it develops more tension in the cold rather than compression in the heat.

 

[hokie66]

Distance between movement joints depends on climatic conditions and structure sensitivity. But for an exposed steel frame with linear members, I wouldn't want to go more than about 300' between joints. I would think the piping would have offsets to allow for the thermal movement, but that is a question for the pipe engineers.

 

[a2mfk]

Jstephen- though off topic, railroad rails have no expansion joints? I can't believe that works, interesting...

 

[JStephen]

In the old days, the rails were 40' long bolted sections and made the nice clickety clack sounds. Now they're continous welded rail, I think shipped in 400' pieces or something of the sort, and then welded. Basically, they can develop sufficient axial tension and compression capacity to deal with thermal stresses. I suppose the joints were always the weak points, so it was advantageous to delete them. Occasionally when it is very hot, you'll hear of a rail buckling, similar to having a road buckle in the heat. And the above comes from railfan activities, I don't design tracks or anything.

 

[dhengr]

Sullypape:

If the link JStephen shows is a reasonable representation of the system, 120' high for the pipe seems mighty high to me, apparently a mirror movement, sun tracking and focus issue. I know nothing about the design of the actual boiler system and the pipe, so this is all imagineering. The piping people will have to provide movement joints where the boiler pipe goes from horiz. (up high) to vert. and down to the ground, not your problem, except as this may affect the last structural A- frame, and apply any secondary loads to it. As long as the pipe is not restrained lengthwise it will just expand, but it is your primary structural (stability) element along its length, so it may have to be braced against buckling btwn. the A-frames. The piping people will tell you what the max. A-frame spacing is. What loads are you having to design around?

Assuming the pipe is supported and braced at mid length, and unrestrained w.r.t thermal expansion lengthwise, you have (.0000065 for steel pipe or .0000099 for stainless stl.)(250° - 0°)(900') = 1.5 to 2.25' of expansion over 900'. The A-frames must be hinged at their base, and as they support the pipe saddle (not roller saddles), so they can move with the pipe as it expands; obviously the first few A-frames, near the mid-length restraint structure won’t move much laterally, but the end most A-frame will need a hinge base and top which will allow btwn. 18 & 27" of lateral movement at their top. As I moved out, on the length of the pipe, I would set the A-frames to be vert. at some mid-temp. range, so they only leaned laterally 9" to 14" either direction from plumb. I would hang the pipe saddle system from the top of the A-frame to allow the hinge rotation at the top, and all the A-frames would be detailed the same up at the top. The base plates at the two legs of each A-frame may be rockers, maybe as crude as a piece of extra heavy pipe, welded to each leg, in the plane of each A-frame, and strapped down to a base pl. and the found., akin to two pipe straps and 4 A.B’s., to allow the rocking of the A-frames. A better detail might be a half-pipe, concave down, welded to the legs, on a second pipe which is the main base piece at the found. This tends to be self cleaning. There are some good plastic and reinforced Teflon type bearing mat’ls. which could be bonded to the two different pipe pieces which would make this joint almost frictionless. The main boiler pipe must take the sum of the force components from the leaning A-frames, plus its own bending and lateral stresses.

 

[sullypape]

To further clarify:

First, I am a structural engineer, responsible for the foundation system for various sites around the world for this system with various modifications, and am trying to confirm the loads from the boiler receiver/reflector support structure and loads from the boiler tubes and receiver enclosure.

As a structural engineer with 30+ years of experience, I have always seen limits of approximately 400 feet for steel pipe rack structures, ductwork structures, and buildings between expansion joints.

I am willing and able to design foundations for a longer structure without expansion joints but then base rollers along the lines of what dhengr describes or sliding joints for the support interface in the reciever at the top with struts and cable braces would be required.

The vendor/product engineers responsible for the superstructure and boiler tubes and receiver believes the system as designed with the A-frames being pushed at the top except for the central 3 A-frames deflect. The centRAL 3 A-FRAMES are braced with tension cables.

Any other comments, thoughts, and/or references would be greatly appreciated.

Thanks,

Sullypape