flowplates / dial-a-pipes

[mcecasf]

Hi there,

I'm trying to find out if there is any guidance - regarding when it is would be recommended NOT to use a flowplate / dial-a-pipe type arrangement for routing flow.

I have a flowplate in a food factory that has 4" piping and line pressures of >8 bar during the cleaning cycle (of 80 degC 2% caustic soda). It had been installed (before my time) using RJT unions on the plate and fittings which - after a few near miss incidents of loss of containment with people nearby - we have changed for DIN fittings that can take the higher pressure. In an ideal world we'd install an automated routing valve block, the problem being that the cost of this - and the associated works around automation, creating the space for the routing matrix, etc - is relatively substantial and there is no real financial payback. One assessment would be that the improved fittings, plus regular maintenance / renewal, should be enough - after all, if you already had a bit of kit to withstand pressure X, why would you then spend even more money replacing it with a fancier solution for a much higher cost if there were no additional benefits?

I guess it comes down to our interpretation of the risk involved in the constant (~12-15 operations/day) removing and resiting of the swingbend and the likelhood of it being connected properly, but it would be nice to know if there was any specific guidance or justification from a safe design perspective..... either to give a technical justification for installation of teh mixproof matrix or to soundly bury the idea.

Grateful for any help!

 

[Latexman]

Personally, I do not understand what you mean by " flowplate / dial-a-pipe arrangement ". Maybe it's local terminology and a sketch or picture would help us / me understand.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[mcecasf]

Hi Latexman,

Routing manifold for transferring fluids between different locations using a swing bend or flexible hose

 

[Latexman]

If the swingbends have to be manually installed each time, automated valves make little sense except from a safety standpoint, because the operator has to be there anyway. Look to the plant's history and the industry's history with these things and compare that to the acceptable risk the company is willing to take on.

With food and hot 2% caustic, it seems that with proper PPE, operating discipline (proper fit up, slowly open valves), secondary monitoring, etc. would be acceptable to the practices I see in the chemical industry.

The justification for automation has generally fallen into the categories of productivity, reduced cost (reduced staffing required), and improved quality.

I see exactly what you mean with your application; I don't see much justification.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[LittleInch]

I'm sorry but the phrase "no real financial payback" compared to "a few near miss incidents" and 12 to 15 operations a day is close to criminal liability. How are you going to feel when someone is blinded by being sprayed with this stuff because of this lethal set up. I've not seen such a high potential for getting it wrong for a long time and it is recognised in the petro chem industry that making and breaking connections is a key area for safety.

This is, IMHO, a recipe for disaster. If those valves we can see are the only isolation between a human and some noxious chemical then this is a very bad solution.

I can't work out if you're trying to justify it or you have concerns yourself. If I was you I would write a clear message to your superiors staying the safety concerns, take a print of the message and keep it in your back pocket for when the safety authority come to call when someone is seriously injured in this disaster waiting to happen.



My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Latexman]

I inadvertantly left out another common justification for automation - safety.

Is there a PHA on this operation? Probably not, 2% caustic is not a listed PSM chemical. But, if there is, take a look at the PHA on this operation. Has a LOPA or Quantitative Risk Analysis been done? Do current operations meet corporate acceptable risk criteria or are there gaps? It should not be difficult to get a somewhat non-subjective determination. If there is no PHA or LOPA or QRA, can you get someone to do one for you accoding to corporate guidelines? It's about the only way to take the personal bias out of the equation. Everyone is biased by their personal experiences and feelings. I know I've been biased over the years

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[mcecasf]

hi guys, thanks for taking the time to comment.

Latexman - the proposed solution would eliminate the plate altogether and use routing valves so the person would not have to be anywhere near - and there would be no swingbends involved. Specific PHA / PSM requirements (I had to google) are not relevant to the UK, but "suitable and sufficient risk assessment" is relevant for our relatively small, relatively low hazard site - in most instances we use HAZOP or basic "likelihood x severity = risk rating" type assessments. In some ways if we hadn't done the initial work to uprate the unions on the flowplate I would have a better time justifying the bigger works as I could definitevely say "the flowplate is not fit for purpose - fittings are under rated for pressure - and there have been X incidents in Y months", whereas now we have an arguably fit for purpose installation so a lower likelihood of failure so what justification is there to change?

Littleinch - The issues that gave rise to this proposal were not from "flow going the wrong way" / "the swingbend being in the wrong position" / "the wrong valve being opened"........ but rather "the swingbend not being perfectly connected, and liquid spray from the union" / "the swingbend blowing off under pressure". My view is that the best technical solution would be to do the work and get rid of the flowplate, but I have to convince the powers that be that the £300k spend (when all associated work, disruption, automation, etc is taken into account) is "reasonably practicable" with respect to the real level of risk from the new, higher pressure rated setup. Hundreds of food, dairy, brewery and smaller volume tertiary pharma sites happily use this kind of arrangement for flow and CIP routing (possibly others too, but that is where my experience stops ;o). Often there is a proximity sensor installed for each swingbend location, so that flow cannot be directed down a route unless the swingbend is in the correct position - so there is control there to stop the operator just opening a valve and getting soaked in something nasty. (Ours does not have a proxi sensor, and in fact implementing such a control would be very difficult with this particular unit and this thought may just given me a useful angle in my justification....). We have multiple plates like this, but generally operating at lower temperature cleaning / less aggressive chemical and lower delivery pressure - the specific one I am looking at operates at what would (to many of this sector) be considered "high pressure" (8 bar is "high" if what you are used to is your high pressure steam supply being 4 bar....).

Bearing this in mind, and picking up a thread of thought from your comments, does anyone know of somewhere on the web I could get reference data for failure rates on making / breaking different types of connections (ideally hygienic screwed connections)?

Thanks in anticipation!