8 replies to this topic

[halflife9]

I'm looking for feedback / experience with padding vessels in propane service for additional NPSHa when pumping at higher rates. I need an additional 5-6 feet of NPSHa for 850 GPM propane import feed rate on a Bingham 3x6x14 horzontal centrifugal pump. Is it possible to achieve this by padding the vessel with hydrogen? I am trying to decide if this is a valid option for the NPSH problem I have.

The current pump is rated for 390 GPM, but could reach 850 GPM with a larger impeller and a total of 17 feet NPSHr according to the pump curve (this includes a 2 foot safety margin). The bullets have a 19.25 foot elevation from the 0% level, but we have a minimum level of 45% (12 foot span) that yields a total minimum NPSHa of 24.65 feet. The estimated line losses at 850 GPM are ~1 PSI leading me to believe the pumps would cavitate at the high rates.

The current system is a set of two propane bullets that vary in vapor pressure with changes in ambient temperature conditions. The bullet pressure ranges from 90 - 200 PSIG throughout the year and the system is rated for 275 PSIG.

Are there any issues with trying to run the vessel pressure higher with a hydrogen gas (95 mol% hydrogen, 5 mol% methane) to "pseudo subcool" the liquid propane feed for additional NPSHa? This could be a cheap alternative to adding a third pump or increasing the elevation of the bullets to increase the capacity. I have considered modifying the size of the common suction header from 8" to 10", but the upgrade will only give me an additional 2-3 feet of NPSHa.

I was hoping there might be some experience available if this has been successfully accomplished in the industry. Contamination of the propane feed with hydrogen isn't an issues, but cavitation of the pumps at high rates is not acceptable.

I can't use fuel gas or nitrogen due to contaminants to the feed system, but these were also considered. I am having a difficult time picturing how the hydrogen gas pad would work in this application. I imagine the amount of hydrogen in the propane should be insignificant at the pump suction, but any feedback here would be appreciated.

Thank you

[Art Montemayor]

Yes, you can blanket your Propane tanks with Hydrogen gas and obtain a higher available NPSH. You are simply increasing the source system pressure when you do this. I have done this very thing various times with nitrogen and even CO2 when dealing with other fuel liquids. I've used nitrogen with LPG as well.

What you haven't contemplated (or mentioned) is the fact that you will get some solubility of the Hydrogen into the saturated liquid Propane and this will exit the tank and have to be made up as time goes on. Any gas is going to exhibit some solubility into the saturated liquid Propane. It's a matter of whether you can tolerate or accept this fact. Being the smallest molecule in the Universe, Hydrogen may exhibit more solubility than other gases.

Having a 1.0 psi pressure drop in the saturated liquid Propane suction line to your pump is hard to accept. I challenge that figure and would request a copy of the pressure drop calculations. I would not accept that high a pressure drop in pump suction lines – particularly that of a saturated liquid system. If you are going to increase the suction line size, how did you arrive at the size? Did you do a detailed pressure drop calculation?

Whether your Bingham cavitates or not is a matter of the NPSHa and the suction line pressure drop – all the way into the eye of the impeller. That is why I would try to get the suction line as big as I could justify – to decrease the pressure drop and to bring down the velocity (& Reynolds Number).

How much Hydrogen will dissolve into the saturated liquid Propane depends on the temperature, the pressure, and the purity of the Hydrogen. I would try it and see what the end results are.

[Zauberberg]

Also make sure that your pump motor can take the amp load with the new impeller/capacity. Centrifugal pump motors usually should be selected based on the largest impeller, however this is not always the case.

[halflife9]

A detailed pressure drop calculation based on the following:

200,000 pph liquid propane (95 mol% propane, 5 mol% ethane) at 90 deg F and 188 PSIA

Flow is parallel through identical piping runs (one for each bullet) 100,000 pph each

65 feet of sch 40 SS 8 inch pipe (Leave as is)
Seven 90 degree elbows
2 Gate valves
1 Motor Control Isolation Trip Valve
1 Welded Tee
Elevation at 19 feet 4 inches
Delta P = 0.13 PSI

Flow is combined through the following pipe run for 200,000 pph as follows:

72 feet of sch 40 SS 8 inch pipe (Upgrade to 10")
Five 90 degree elbows
1 Welded Tee
1 Gate Valve
8 inch x 6 inch pipe size reduction
Elevation at 13 feet 4 inches
Delta P = 0.27 PSI

5 feet of sch 40 SS 6 inch pipe (Upgrade to 8")
Two 90 degree elbows
1 Gate Valve
Elevation at 0 feet (pump centerline)
Delta P = 0.48 PSI
Total DP = 0.89 PSI at high rates excluding the Suction Strainer (6" Conical strainer with Mesh Screen)
The strainer has an inside screen with a 0.063 inch (14 mesh) / outside screen 0.018 inch (4 mesh)
It says that it is only rated for 360 GPM (I am sure this could be an issue).

The 0.89 PSI is about a 4.6 foot NPSAa loss, which is reduced by 56% to 2 feet upgrading to 10 and 8 inch pipe runs. This would barely get the job done at 17.33 NPSHa vs 17 NPSHr, but it leaves no rooom for error and the pump suction nozzle is limiting the pipe at 6" anyway. I am sure the suction strainer has a significant Delta P, but I don't know what to use in the calcualtions.

[djack77494]

halflife,
My suggestion would be to evaluate replacement of your pumps with new ones able to operate with low NPSHa. There are various manufacturers of such pumps, specializing in pumps suitable for liquified gas services.

Let's talk things through. You are wanting to more than double their current capacity (390 --> 850 gpm), which seems like a pretty tall order. And you're thinking you can do that with simply an impeller change. Most unlikely under normal circumstances. Add that to your suction lines have more than 140 feet of straight pipe. Like Art, I was prone to question the suction pipe losses, but I didn't imagine this run length. I do not think your suction strainer would have much loss, but do feel that you could remove it altogether after commissioning. I'm still not fully understanding your figures. If you start with a minimum liquid level that is 19.25 + 45%*12 = 24.65 ft above grade, and lose 4.6 ft on the way to the pump, you'll still have 20 ft of NPSHa, no? So you'd have a margin of 3 ft. If that's the elevation of your pump centerline, then things make sense.

I abhor the idea of adding a blanket gas to increase system pressure. You may consume significant amounts of hydrogen, which I think would make the system's designer not look good (or worse). You'd be increasing the vessel's pressure, but it is well designed for holding liquified LPG (and nothing else). You'd need hydrogen bottles and a pressure regulating system. Messy. I'd rather buy new pumps, raise the vessel's operating level (even at the cost of lost storage capacity), increase the size (and reduce the length) of the suction piping, or even raise the elevation of the bullets if needed. Of course I understand that those options are more costly, but I couldn't be comfortable with this low cost option.

[JoeWong]

Significant consumption of H2...
Contamination of propane product...

Have you checked...

i) if the associated electrical devices is compatible with present of hydrogen...
ii) if the product contaminated with H2 and acceptable to storage tank on the receiving facilities...
iii) if storage metallurgy is fine with present of H2 (HIC, hydrogen embrittlement...)
iv) reliability of H2 supply...

[halflife9]

Djack77494,

My numbers were misleading since I already included the 5.4 feet from the minimum liquid level as part of the 19.25 feet of NPSHa. The actual elevation from 0% level was 13.8 feet.

I appreciate all of the responses and I agree that this system is probably being pushed well beyond its limitations. Unfortunately it is sometimes difficult to convince others that what appears to be the cheapest solution isn't always the way to go.

I am hoping that a combination of the upgraded suciton piping / impeller modifcations will do the trick. I am still waiting to hear back from some vendors, but the hydrogen pad idea is still lingering in the background.

[benoyjohn]

Hi halflife9,

If you are going for new pumps you may check out the option of 'can' pumps.

regards,

Benoy

[JoeWong]

I have similar experiences...Just make sure those safety issue (as highlighted earlier) which can bring catastrophe are addressed.