Internal floating roof tanks level switch installation

[danschwind]

Dear colleagues,

One of my clients asked me if I knew any method to install a level switch (for overfill protection) in vertical atmospheric tanks with internal floating roofs while the tank is in operation.

The tanks would be mainly storing gasoline (although some may be operating with other fuel, like ethanol or diesel, during the installation of the level switch), 6m to 19m diameter, 7m to 17m height.

In an ideal scenario, a regular level switch (optical, thermical, vibrating fork, etc) would be installed inside a stilling well that would go through the floating roof or would be installed in a external pipe hooked up in the tank's wall. The problem that my client faces is that both these scenarios would require the removal of the tank from operation. This wouldn't be a big deal if it were 1 tank, but they would be performing this nation-wide (not USA, as it can be inferred from my not-so-great english) and the number of tanks would be dozens.

Has any of you dealt with a similar situation? Is there any technology or installation method that would allow the installation of a level switch for overfill protection without removing the tank from operation? I honestly don't think there is a solution for this, but if anyone knows, he is probably here in Eng-Tips

It can be a premise that at least 1 nozzle at the tank's fixed roof (on top of the floating one, obviously) is available.

Thank you all in advance!

 

[IRstuff]

What entails "operation?" Tanks fill and drain, so what frequency, duty cycle, etc., while under "operation?"

How does the floating roof float? Is it just a bladder kind of thing, or is it rigid, or what?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[danschwind]

Dear IRstuff,

I'm sorry for not defining operation.

Operation = actively storing fluid, the oposite of being totally empty and inert (without an explosive atmosphere), that allows hot-work or internal access to it.

I don't think it would be the end of the world for my client to isolate the tank from the inlet and oulet lines for the duration of the installation, as the real costly procedure would be to empty the tank and to make it inert (for dozens of tanks). Therefore, if it helps, the liquid level may be considered fixed.

I'm not an expert at floating roofs (in portuguese we actually differ floating roofs from what I'm actually trying to express, that in a free-translation would be called "floating seal"), but in overall it is a very thin sheet of aluminum that sits on top of the liquid. I would then say it is kind of a bladder thing.

I do have some drawings of an example that I could attach, but please note that the drawings are in portuguese. Let me know if you need them!

 

[IRstuff]

OK, so if the aluminum is reliably floating on the surface of the liquid, then it could be considered to be the surface of the liquid, in which case you could wait until the tank level is sufficienty low and install an ultrasonic or radar level sensor on the actual roof.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[itsmoked]

That's what I'd do too.

Face it, the time and hassle to install these eclipses the cost of the sensor or switch.

I'd NOT use a 'dumb' switch but would instead use a radar distance sensor. This then supplies accurate backup to whatever you're using now for the tank fill level - easily the most important variable in tankage.

Switches are also lame in that they are often -especially in your application- hard to test and prove operational. Not good for the critical protection needed against overfill. With a sensor instead of the switch you get constant and daily feedback of its operation showing that it's functioning. Something completely lacking in switches.

If you have an existing tank level system you also get the added valuable security of having it to compare against each other. I'd even log that so if the two start to diverge the operators are alerted and looking back at the log can show when one of them started to fail.

You use a readout with a level setting alarm output to do the 'switch thing' you may need with your system. I'd probably include a colored stack-light

that shows green/amber/red to telegraph to everyone how close the system is getting to overfill.

An example (Radar):Rosemount

Recently LIDAR has stormed onto the scene too.

And, your English rocks. Don't sell yourself short, you could probably teach English in the US.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[danschwind]

Thank you all for the replies!

As a matter of fact, some of the tanks will receive radar systems, but not the ones that we are dealing here.

The ones that are the focus of my questions are the ones that nowadays have absolutely zero overfill protection. Those tanks operate in "second-tier" storage bases usually in low-demand regions. The level is measured manually (through a tape measure) and its "continuous measurement" is updated by the inlet vs outlet readings. The project in question is to upgrade those tanks to have at least 1 level switch intended for overfill protection.

Installing radar systems in those tanks are, unfortunately, out of question due to cost issues. Also, it would be hard for me to convince them to use a radar not for inventory measurement but only for an overfill protection (trust me...).

An ultrasonic level sensor could theoreticaly work in those scenarios, by measuring the ullage down to the floating roof. In my understanding they are also cheaper than a radar, although probably still way more expensive than a regular level switch. This difference in value could maybe be offset by allowing the installation without all the procedures needed to empty and inert the tank. I will study more about it, but it may be a viable choice

If you guys have any more ideas, they are welcomed!


itsmoked, thank you for the compliment!

 

[danw2]

Top mounted, custom probe length Siemens CLS200, a capacitance-technology switch, can detect levels of liquids with dielectric constants above 1.5.

Also provides some diversity from the radar tank level for inventory.

 

[thebard3]

Our standard tank level protection requires redundant level monitoring devices. Most tanks have an active analog level sensor (we have recently standardized on guided wave radar) with a secodary device being a discrete level sensor. We have subsequently found that ultrasonic sensors are just about the same price as the discrete devices so we are now using a combination of GWR and ultrasonic. Also, we monitor any deviation between the two devices as an alarm condition so that provides an accuracy check for failed sensor. As your tank top is not a fixed point, how do you determine the product level from above? Are you looking at the actual height of the tank top? Perhaps a device in a standpipe would be better, but this would obviously require the tank to be removed from service for installation of the standpipe.

Brad Waybright

It's all okay as long as it's okay.

 

[danschwind]

Yes, theoretically the ultrasonic sensor would measure the height of the floating roof and not the height of the fluid itself. As this is intented for overfill protection only, they need reliability, but not so much precision (in my understanding at least).

And yes, using a standing pipe would be better. If no cheaper alternative is found, they will install in a standpipe or in a lateral pipe device, but I can't begin to fathom how much money they would lose to perform these operations.

 

[Compositepro]

I think that one of the most common level gauges for large tanks is still the old weight on a cable that goes over a pulley at the top of the tank. Applying a magnetic switch to the weight on the outside of the tank is quite simple. If a heavier weight is placed on the cable end inside the tank, you do not even have to enter the tank to attach the cable to the floating roof.

 

[itsmoked]

Whatever solution you use you probably need either explosion-proof or more likely intrinsically safe hardware.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[georgeverghese]

In larger companies, it is customary practice to run a formal SIL study to determine the number of transmitters to be used at these tanks,and the results may vary depending on the flammability of the service fluid, whether there is sufficient containment volume in the dykes, proximity of the tank to nearby civilian infrastructure (housing, hospitals etc), level of process automation and if the control room is manned during fill operations.

For example, in some critical flammable applications, the results may indicate one control transmitter to cover the entire level range (with high level alarm at CR), and a 2oo2 voting arrangement for SIL 1 transmitters in high level trip function to act on 2 automated shutdown valves in the feed line.

Ask the process safety engineers to rank these tanks according to risk and find an external contractor engineer to run these SIL studies for you. The actual mechanical details of installation can be addressed later.

 

[danschwind]

Dear georgeverghese,

Those studies have (theoreticaly) already being made. They already have the approval to proceed with the installation of those level switches on tanks that currently don't have any. Their problem right now is exactly the mechanical details of the installation so they asked me if I knew any solution, for which unfortunately I did not. Also, unfortunately, deciding whether this is the best solution from a safety point of view is outside my scope.

They will proceed to install those level switches in non-floating-roof vertical tanks normally, but the ones that do have a floating roof (mainly gasoline ones, so they are numerous) require the removal of the tank from operation, which is a very costly operation.

 

[danschwind]

Dear colleagues,

I have forwarded the idea of an ultrasonic meter measuring the top of the floating roof to my client and it was a well received idea. We will now contact some vendors to make sure the solution works and then proceed to an economical analysis.

I'll update the thread with the outcome, as this may be a recurring problem for another colleague.

Thanks for the ideas!