Tank/Relief Design: Air Surge from Tanker Off-loading

[TiCl4]

I'm specifying an iron (iii) chloride (40%) fiberglass storage tank that is filled by the vendor via tanker truck. The tanker truck typically offloads via air pressure, and has a PSV on the tanker truck at 30 psig.

Typically, the vendor trucker will pressurize to ~15 psig, then shut off air supply at 50-70% offload to let the tanker depressurize as liquid level in the tanker truck drops. The over-pressure scenario is if the vendor trucker does not draw down air pressure during offloading, and is pressurized to the maximum allowable when all liquid off-loaded. This will lead to a scenario whereby an large air-surge will go through the pipe and into the tank, necessitating a decent vent size to handle the surge.

FYI we have split other fiberglass tanks nearly in half in the past due to this issue - they were not appropriately sized for the air surge.

I'm in the process of specifying the tank to the vendor for quote. The tank is going to be a fiberglass tank, likely Derakane 411 or similar. I have a few questions/concerns about the overall system design:

1. What is the maximum economical pressure rating for a fiberglass tank (especially a flat-bottom/dome top storage tank?) It is likely cheaper to specify a larger vent than a thicker tank, but I'm not familiar with fiberglass tank pressure rating economics.

2. The current tank to be replaced is bottom entry. Iron chloride has a tendency to foam, and I am concerned that the relief of the 700-1000 SCFM (my initial swag, not detailed calc, ~100 ft of 2" fill pipe) may become two-phase at such a large air flow rate. I'm thinking I should design the system for top entry, but FeCl3 will foam on splash entry. That is not desirable either. Directing liquid flow towards the side of the tank would reduce foaming, but I'm concerned that the solids in the FeCl3 liquid may abrade the surface of the tank if I direct flow towards the side of the tank. Perhaps the best solution would be a bottom feed with a hinged manway (24", perhaps) for relief, accepting that relieving conditions may be two-phase? I'm not sure the best direction to take on this. Help would be appreciated for both tank and relief design options.

 

[don1980]

In my experience, the design pressure for plastic flat-bottom tanks is typically zero (atm). In some cases you can find flat bottom plastic tanks designed for a few inches WC, but rarely more than that.

Alternatively, the you can specify an FRP pressure vessel, designed per ASME Sec X. These are becoming relatively common. I've seen Sec X FRP pressures as high as 90 psig. As you would expect, these are much more expensive than flat-bottom FRP tanks.

For the lesser expensive option (flat-bottom tank), a large atm vent pipe would be my first choice, assuming that an atm opening is acceptable for this service. The possibility of blow-through from a pressurized tank trailer is definitely a scenario for which protection is needed. Depending on the size of the tank, it may not be practical to install a sufficiently sized atm opening. If so, I'd install an auto shut-off valve on the inlet, activated by a pressure transmitter on the tank. Personally, I'd design this to be a HIPPS, for the higher integrity. If you do this, determine the worst case amount of vapor space (smallest vapor space) in the tank because this affects the amount of time you have to shut off the incoming gas. The smaller the vapor space, the less time you have to shut off that flow.

I suggest you first contact an ASME Sec X FRP pressure vessel fabricator to get a cost estimate. If it's affordable, that's the safest design.

 

[georgeverghese]

Agreed, install a PSHH on the FRP tank to auto shutoff an inlet valve to address this gas blowby scenario. You can limit the emergency inflow by installing a restriction orifice on the feedline to the FRP tank. Choose the FRP tank pressure such that the gas blowby relief rate (as limited by the RO) is handled by a reasonable size breather / emergency vent valve on the FRP tank. Calculate the gas blowby rate based on a source pressure of 30psig at the trailer tank.
Is there some way you can get a feedback signal from the trailer truck tank level? Shutting off inflow to the FRP tank based on level on this trailer tank would be better than to wait for the PSHH on the FRP tank to activate.

 

[TiCl4]

I find it a bit confusing when people say the tank is designed for no pressure. When relieving anything to atmosphere, a dP must exist. Therefore even an atmospheric tank experiences pressures above atmospheric in order to breath. There must always be a design pressure, even if it is 10” wc like for polyethylene tanks. Is 0.5 psi MAWP on a small fiberglass tank really so difficult to achieve?

This site does not have the safety culture for a HIPPS to be a reasonable option. Even PSVs are not routinely checked here (meaning they are never checked)...I need as passive of a system as possible that doesn’t rely on routine function checks, which is why I was leaning towards a “weighted” manway at the top.

Are you suggesting a PSHH shutoff in addition to an appropriately sized relief? Running a mini LOPA here - Initiating frequency of this event would be 0.1 due to the frequency of unloads (10/yr)/standard operator error (1/100 error rate). Properly designed relief is another .01 (two orders of magnitude), leading to baseline frequency to be 1/1,000. That would mean another 0.01 is required, which probably should the the PSHH shut off at a SIL-2 rating. I can’t take credit for no personnel in the area - it’s at least a 50/50.
Our site is not at a level to keep up with SIL requirements...any other suggestions on additional measures? It’s a no-go on the vendor’s truck level feedback, btw. Your suggestion of a pressure-rated tank might be necessary here to avoid the overpressure scenario in the first place...
Thanks for the input so far, all. I had no idea fiberglass pressure vessels existed.

 

[LittleInch]

"Atmospheric" tank design is I agree a little confusing, but is commonly understood by vendors.

As the weight of the product and FECL is an SG of 2.8?? creates vastly more pressure at the base than a few inches of water column pressure.

~API 650 which is for large steel tanks defines atmospheric as <= 4 inches of water column or 1kPa.

The impact of air pressure goes up with size so without knowing how big your tank is it is difficult to say, but 10 inches / 0.5psi / 2.5 kPa is probably as high as you want to go.

Your issue I think is more the speed of opening of any vent / pressure relief. A bubble of 30psi air could create quite a significant over pressure for a fraction of a second before the manhole lifts.

If you want something simple and relatively bullet proof then I would recommend you have an intermediate larger pipe, say 12" or 16" built to withstand pressure and place your relieving device / air eliminator on that with a non return valve on the exit to the tank. You would need to set it to relieve above the static head of the tank or lift this pipe up to the top of the tank and allow the material to drain down into the base of the tank. What ever you do you will need to be careful about emitting foam or liquid spray with the air.

For the tank situation you need to think about the max head in the tank at the end of loading with respect to air volume in the tank and distance from liquid level to your vent.

FRP pressure vessels do exist, but they tend to be fairly limited in diameter and are also quite costly compared to more simple vessels.

FRP is notorious for not liking shocks and damage of any kind so you need to be extra careful for these.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MJCronin]

Consider a steel tank with a fiberglass (or other) liner...

Things are out of control in the bulk chemicals marketplace .... Delivery drivers want to offload as quickly as possible and get back on the road.

The tank solution that you select today, may not be acceptable with the cowboys of tomorrow ....`

MJCronin
Sr. Process Engineer

 

[Latexman]

More passive. Use a pump; not compressed air.

Good Luck,
Latexman

 

[TiCl4]

Product sg is 1.4, not 2.8 (according to vendors sds and Solvay’s charts. The tank is only a 7,500 gallon tank.

Pump offloading might not be an option. Currently the vendor hooks up and offloads their own tanker - I’ll check with the supplier. Also, iron chloride basically requires CPVC, fiberglass, rubber, or titanium wetted parts. Diaphragm pumps are not good on plastic piping, and I’ve not heard good things about plastic centrifugals. How are fully rubber-lined centrifugals? Seals would be an issue here, and mag-drives are likely not good due to the particulates in the FeCl3. I don’t think the powers that be will let me shell out the $$$ for a titanium pump.

I’ll look into FRP-lined or rubber-lined tanks. Anyone have experience on rubber-lined tank longevity?

The stand-pipe on the loading line is an option too. I’d probably still have to put a hinged lid on it to prevent rain intrusion as well as siphoning air into the feed stream if it is on a top entry.

Lots of great responses, all, thank you!
Thanks for the great suggestions, all

 

[Latexman]

So, you gravity transfer out of the tank? I figured you already had a pump that works there. One pump that offloads and transfers? Can you increase productivity by speeding up the transfer to justify a pump?

I'm pretty sure a unit I worked at 15 years ago had a iron (iii) chloride solution pump. If you Google "iron (iii) chloride solution pump" there's several vendors that say they make iron (iii) chloride solution pumps.

Good Luck,
Latexman

 

[TiCl4]

Latexman - yes, we gravity feed out of the tank into treatment basins. There are no pumps in the system, currently. This is used in a wastewater application, and we get maybe 1 tanker a month, which does not affect production in the slightest.

 

[georgeverghese]

Standard process safety design practice is to install both an instrumented shutdown mechanism and a fully sized relief device in the event of failure of operating procedures and coincident failure of the instrumented shutdown system. Frankly I am not a fan of these failure frequency stats; all I can say is the probability of a relief event here is higher here than in standard continuous process systems due to the higher reliance on operator intervention (as compared to the BPCS for a regular plant process unit operation). Made worse by the fact that these "operators" are actually untrained truck drivers?

 

[KevinNZ]

The tank must have a vent system to allow it to fill. Why not just a larger vent pipe?

 

[Iomcube]

TiCl4, we always used large FRP tanks for pressurized gases (Cl2) of no more than 1barg! For hypochlorite & brine solutions I must say that longevity of FRP lines CS vessels far exceeded than rubber lined solutions. In no FRP vessel (which we primarily used as buffers to avoid gas surges & aid our instrumentation) I had seen PSV. Designs of these were vetted by UHDE-GMBH