Diesel Auto Ignition Pressure 2

[Brent Boisvenue]

I'm designing a fuel distribution system for a mine with fuel piping going down a 600 m (~2000 ft) mine shaft and I'm trying to find some relevant literature on diesel auto ignition PRESSURES. I'm finding lots of information on diesel ignition pressures in internal combustion engines, but mainly regarding transient properties and ignition delays related to pressures.

My application is very simple. I need to figure out how much static pressure in a pipe would bring about a risk of the diesel fluid igniting. I've seen 80 psi, but I have nothing to substantiate it.

Any petroleum engineers out there with some pointers?

Thanks!

 

[TugboatEng]

No risk.

We have pipes operating at 37,000 psi with diesel. You might want to fill the pipe slowly to avoid creating heat from compression but once the air is purged the fuel can't burn.

 

[BrianPetersen]

It's not the pressure. It's the temperature ... and being mixed with air ... which, in a pipe containing only fuel, it won't be.

Complete non-issue.

 

[Brent Boisvenue]

I should clarify that there will be air entrained in the pipe. Fundamentally, we are designing the system to provide open channel flow. So, air will certainly be an issue with this application. We have a pressure transmitter at the bottom of the pipe run that will signal the pump on the surface tank to stop if it reaches a certain threshold in the event that the flow fully develops for some reason.

 

[BrianPetersen]

That's more complicated, but still ... it's not the pressure, it's the temperature.

Compression-ignition in that situation seems unforeseeable, but if standard diesel fuel and air are involved, it is conceivable that it could reach the flash-point temperature, and be vulnerable to ignition from some other source (e.g. electrostatic discharge).

The ratio of the air and fuel have to be within the ignition / explosivity limits for anything to happen at all. If the temperature is above the flash point ... that's not impossible.

I would respectfully suggest to try to avoid having air and fuel mixed, because if they're not mixed, you know that there will not be a problem.

 

[Brent Boisvenue]

Thanks Brian,

This is an interesting discussion. The design I'm working on is an extension to a previous project that was executed a few years ago, where the system was specifically designed to meter the diesel into the underground fuel tank from surface using flow control valves to keep open channel flow in the pipe. The system had a PIT installed at the low point of the piping interlocked with the fuel pump on surface that would shut down the flow if it ever reached 80 psi as a safety measure. This is where the 80 psi parameter comes from that I can't validate. I can't validate whether reaching 80 psi is a conservative upper limit or still within a hazardous regime for air-entrained diesel. It could also be that 80 psi was used as an upper value used in the piping strength design. As far as I understand, this system has been working well for our client, but that's not to say the 80 psi upper limit is not on faulty premise, especially if fully developed flow has never been reached and the interlock has never ben used (tested in real life).

Not having worked on fuel systems before my first instinct was to match the operating philosophy on the existing installation, given that we are filling from the same tanks on surface. However, my application is fundamentally different. The existing installation has the fuel piping running down the mine shaft and I'm guessing they didn't want to keep the piping filled in the shaft to avoid the risk of leaks and combustion where the mine conveyances are running (including the cage for mine personnel). The total differential height from the surface batch tank to the underground tank was over 600m. The existing system includes several vacuum breaks and seems to be designed specifically to ensure open channel flow, which would guarantee air entrainment.

My application has fuel delivered directly to the fueling station underground via a dedicated line in a bore hole (hole drilled in the rock) from the surface batch tank and I only have a 95m differential elevation (115 psi of diesel static pressure). So, keeping the line charged with fuel would be less hazardous.

NFPA 122, Section 8.3.4 also has provisions for "wet type" or "dry type" transfer pipelines. So, if I'm interpreting this correctly, it seems there is some precedent for both scenarios. It may be simpler in my case to keep the line charged and avoid air entrainment. I'd have roughly 300L of fuel charged in the line above the tank, but it's a smaller volume than the 5,000L tank I'm filling.

 

[SwinnyGG]

Are you absolutely certain that this 80 psi limit was previously set based on some perceived autoignition danger (which, full disclosure, is nonsensical) and that the 80 psi limit is not related to the pressure tolerance of the pipe in the system?

This sounds to me like a failsafe against overpressure in the pipe, not a failsafe against autoignition.

For reference the flash point of diesel fuel is roughly 125 F, and the autoignition temperature is roughly 400 F. Flash point is the lowest temperature at which a combustible fuel generates enough vapor for the vapor fraction just above the surface to reach the lower limit of flammability. In other words, below the flash point, the vapor mixture simply cannot ignite.

The other thing to remember about flash point is that the flash point is the direct result of increased evaporation of the fluid due to a temperature increase at standard pressure. If the system is at a pressure greater than 1 atm, the flash point goes up. This means that if the system is at 80psi, the flash point is much higher than under standard conditions and ignition of the vapor is much less likely.

TLDR - as system pressure goes up, for a mixed stream of any flammable liquid and air, vapor pressure goes up and the system actually becomes MORE resistant to ignition, not less, assuming the pressure is held.

If the following three conditions are met, the diesel fuel within your system cannot ignite:

1) No part of the system - ie fittings, sensors, pump parts, whatever exposes a surface with a temperature greater than 125 F to the fuel/air stream. (even though the high pressure flash point of the mixture will be well above 125 F, you need to protect for a leakdown condition back to 1 atm anywhere in the system - so 125 F, or whatever the flashpoint is of your specific fuel, should be your safety limit for surface temps inside the system).

2) The fuel/air mixture never reaches a temperature of 400 F at any point in the closed system.

3) No significant portion of the system is exposed to negative pressures (pressures below 1atm); this would reduce the vapor pressure and cause the flash point to drop.

 

[Brent Boisvenue]

Thanks SwinnyGG,

These are some great insights. I finally found a reference buried into the project files on where this 80 psi figure came from. The previous engineer had found a reference somewhere that says diesel fuel can auto-ignite at 260 psi and he had picked 80 psi arbitrarily with a safety factor of more than 3. However, based on what I've read, that pressure coincides with a finely vaporized air/diesel mixture in the combustion chamber of a diesel internal combustion engine. The physics aren't the same in our application. In an ICE, the pressure is purposely increased for the purpose of combustion. For this to occur in a static column of piping, the only mechanism that could possibly allow this is if there was a trapped vapour pocket entrained at the bottom of the column of piping.

In our application, if the entire column of piping up to the shaft collar were filled with diesel fuel, we could conceivably see up to 727 psi as the column of pipe fills. So, maybe this was his premise. Having designed his system for open channel flow, I suppose he felt there was no reason to design to higher pressure.

Out of curiosity, on your point about a fuel's flash point increasing with system pressure, I've seen references indicating that diesel auto-ignites in a combustion chamber at about 3MPA (~425 psi). So, I'm guessing there's enough air compression to increase the temperature of the air to auto ignite the mixture? Would this not be the case for an entrained bubble of vapour in our scenario?

 

[3DDave]

A glance at several sources indicates the research is entirely about pressures reached with sudden compression to raise the temperature of the air. Your bubble will be entirely cooled by the surrounding fluid and, since there's no other indication, will be compressed only as it goes deeper into the mine, taking several minutes instead of the fraction of a second seen in diesel engines.

 

[TugboatEng]

See common rail injection for diesel engines. There is no issue with pressure. Our MTU 4000 series engines are operating at 37000 psi in the rail. Your concern is unnecessary.

 

[SwinnyGG]

Out of curiosity, on your point about a fuel's flash point increasing with system pressure, I've seen references indicating that diesel auto-ignites in a combustion chamber at about 3MPA (~425 psi). So, I'm guessing there's enough air compression to increase the temperature of the air to auto ignite the mixture? Would this not be the case for an entrained bubble of vapour in our scenario?

The pressure increase in the charge air of a diesel engine happens in milliseconds - which is why there is a very large temperature increase.

Unless you are able to pump at absolutely gargantuan flow rates, as long as the pipe has air in it it is not possible for you to increase the pressure of the line at anywhere near the same rate; this means that the corresponding temperature increase will be much, much lower, even if you reach very high pressures. If your line is uninsulated, as I assume it would be, you will likely get very near zero temperature change in the line due to pressure changes along its length.

Also, in a diesel engine, the fuel being injected into the cylinder is already vaporized; it is not a fluid layer with a vapor layer being generated by temperature change; it is effectively a vapor already. This means combustion is much, much easier.

 

[MintJulep]

You seem to have a very inconsistent set of assumptions.

The same pipe can't be simultaneously a solid column of liquid and "open channel flow".

600 m straight down vertical? Inclined? Stepwise?

What's at the bottom of the pipe?

If the pipe is empty, and you start putting fuel in at the top how much momentum does it have at the bottom?

How does fuel go in at the top?

A slug of liquid descending 600 m will cause a pressure wave in front of it, and a vacuum behind it.

 

[TheTick]

How does one get open channel flow and still be under pressure?

 

[Brent Boisvenue]

Some clarifications: the original system was designed for open channel flow and the flow was limited to that which cannot fully develop in the pipe as long as the line never slopes upward or any obstructions are encountered. There was a failsafe PIT installed at the low point of the piping that was there in the event that something caused a fully developed flow scenario, which would shut down the fuel pump. I don't think this was something that was expected to occur, but the team at the time opted to include it as a conservative safety measure.

Reading through the responses, it seems the 80 psi parameter used as an interlock is on the extreme end of conservative, but in the end, with no obstructions in the pipe leading to the underground fuel tank, I don't think it mattered. The system was installed in 2015 and to our knowledge has never tripped on high pressure.

Thanks for your input everyone!

 

[LittleInch]

That logic is fatally flawed.

The 260 psi relates approx to a compression ratio of about 17, which is what diesel engines run at (between 14 to 25 is what google tells me)

The issue being that the sudden compression that occurs in an engine raises the air temperature to a much higher value than ambient and hence when diesel is sprayed in it ignites.

It's not the pressure, it/s the temperature. Your previous designer should have used a temperature measurement, not a pressure measurement and so should you. IMHO.
You can set fire to diesel at very low pressures but high temperature, especially if you vapourise it. There are plenty of diesel fired space heaters....

I understand the issue about fuel backing up, but if I was you I would use a full pipe in your case as the pipe will be smaller and you can always empty it when not transferring fuel. Then when you need to use it just close the bottom valve, slowly fill it until you get full pipe flow then keep a back pressure on your pipe to avoid going two phase flow and go from there.

Serious mining drops then yes you need open circuit flow or you end up with very high static pressures.

This is one of my archived threads as it was so interesting.

https://www.eng-tips.com/viewthread.cfm?qid=409750

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

 

[MintJulep]

The system was installed in 2015 and to our knowledge has never tripped on high pressure.

Do not confuse 6 years of good luck with a safely designed system.

 

[Brent Boisvenue]

Thanks LittleInch,

I have lots to consider here.

 

[Brent Boisvenue]

I greatly appreciate everyone's input. My interpretation from your responses is that he previous engineer's safety design is extremely conservative. Even if the pressure climbed significantly above 80 psi, the temperature could not climb enough to reach auto ignition. Just to recap, the system flow cannot reach fully developed flow through a control valve and there are no valves downstream of the control valves. The diesel drops vertically down the shaft and into sloped piping leading directly to the underground tank.

My application is tying into the same system downstream of the existing control valves and dropping the fuel piping down a sloped borehole with piping to the underground tank.

I've since managed to touch base with a couple of previous colleagues and have learned that it's apparently commonplace in underground mining to design fuel systems for open channel flow in fill pipes.

So, as long as the line has a flow restriction before discharging into the mine preventing fully developed flow and there are no valves, orifices or dips in the horizontal piping runs, we can guarantee open channel flow and everything is kept at atmospheric pressure. This is backed up by industry precedent.

Is there anything I'm missing? Does anyone see a situation arising that would create a combustion hazard with this system setup?

 

[LittleInch]

Brent,

Your interpretation should be that the previous engineer got it wrong. He should be measuring temperature, not pressure. Pressure on its own won't ignite diesel vapour, high temperature will do that.

Your issue, IMHO, is how to deal with the diesel vapour. When you get liquids falling down vertically or near vertically in a pipe, they bring with them air dragged along with the liquid. Diesel vapour is flammable / explosive. I am guessing the end of the vertical or sloping pipe is either open to atmosphere or has a vacuum valve on it?

If this is a sloped open channel flow, again you will have some air flow if you have an open end into the tank, but if you enter your tank at the base then you might get away with it providing you are operating at a fairly low liquid volume fraction in your pipe. Any idea what it is?

The vertical drop flow it seems is quite normal for mines and I can understand why. If you read the link to my previous post, there seems to be a magic number of 7/24ths for liquid to air volume in the pipe as it falls down the pipe.

The other thing to think about / remember is static electricity. Diesel is notorious for being quite a good insulator and holding static charge from high velocity fluid flow for a long time. So your tank needs to be well earthed and have no metal dip sticks or other items which can create a spark inside the tank. It takes quite a long time for this charge to dissipate into the walls of the tank.

And don't forget the volume of liquid in the pipe when your high level trip goes off in your subterranean tank....

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

 

[Brent Boisvenue]

Hi LittleInch,

I believe you are correct in that the assumptions that the previous engineer used were wrong about the pressure parameter. I don't think it matters in the end. It's an interlock that isn't needed to prevent combustion.

I found a parallel discussion on the same topic here:

https://www.eng-tips.com/viewthread.cfm?qid=81608.

There was no pressure interlock discussed in the thread, but the conclusion as that this is common practice in mining and there is no analytical solution for free-falling diesel fuel in a vertical line.

My case; however is different. I'm delivering fuel in a sloped line and I was able to estimate the velocity using open channel flows. Having never calculated open channel flows, I've also just realized that the Manning approach doesn't take fluid properties into account, so my next task is to figure out how to incorporate Darcy-Weisbach friction factors into account. The velocities I calculated are, of course, based on water and are quite high (~31 ft/s). However, even after I correct for the density/viscosity of diesel, I don't think I'll see any order of magnitude reduction in velocity (I may not go there). Velocity is another issue I'll need to design around. I may need to incorporate a horizontal run underground to transition to fully developed flow to slow it down before entering the tank (also apparently commonplace in underground mining).

On the topic of static charge, my electrical counterpart will be designing for grounding/bonding as required. It's certainly on his radar.

...and I thought this was going to be a boring project.