Heat Transfer in a hot oil heater

[Sam654]

I have a hot oil heater currently running at my max permitted heat input of 165 MMBTU/HR. The production process has fully opened opened all of their users due to fouling issues. Yes, we're aware that they need to clean thier heat exchangers but, we're trying to make it to a scheduled shutdown to do so. The problem is when they opened all their temperature control valves, the Therminol flow through the unit increased from 12,500 GPM to ~13,500 GPM. As a result, I am unable to maintain setpoint temperature.

My explanation to plant management is the residence time through the unit has decreased and since I cannot increase my heat input, my heat transfer to the fluid decreases. They say I'm wrong, the heat load is still the same and the heater should be able to maintain temperature. I have argued back that what they say is partially correct... the heat LOAD is unchanged but, the heat input has dropped due to a decrease in residence time.

Who is correct?

 

[semsagro]

I guess if the oil temperature at 12500 GPM and 13500 GPM is the same, the heat input would increase with increasing flow rate.

 

[MintJulep]

With the same heat input to your heater, if the flow increases the output temperature will go down.

 

[zekeman]

What do you do to maintain temperature? If, as you say the valves are full open, then without changing the input power, you have no control of temperature, the temperature is a unique function of power and flow rate.Fouling of the surface would only increase the metal temperature to accommodate the unique temperature function.

 

[racookpe1978]

Sounds like you need to "re-explain" to them in simpler terms. (Also appears they've run their own plant into the ground and don't want to admit it, nor really fix the problem. or will attempt to fix their own problem with short-term (cheaper) fixes and never address the root cause of the heater problems.) Oh well.

Try this explananiton:

My heaters are maxed out at 165 MMBTU/HR. They can't put any more energy into the heat exchanger. At a lower flow, the oil being heated stayed in the heat exchanger longer, the increased stay time allowed the oil to get to a higher average temperature (higher LMTD) befor eexiting and going to your downstream processes.

Now, with flow max'ed out, the oil doesn't stay in the HX long enough to heat up as much as it did before.

---...---

Can you hire a temp services truck as an auxiliary heater for a few weeks through the winter season. Put it in parallel with the original HX, or into the storage tank for the oil so the arrival temps of the oil going in the HX are greater.

 

[zekeman]

"My explanation to plant management is the residence time through the unit has decreased and since I cannot increase my heat input, my heat transfer to the fluid decreases. They say I'm wrong, the heat load is still the same and the heater should be able to maintain temperature. I have argued back that what they say is partially correct... the heat LOAD is unchanged but, the heat input has dropped due to a decrease in residence time."

No, see MInt Julep's explanation. Put another way:

Q=M*c*(To-Ti)

To temperature leaving
Ti themperature entering
Q = heat input fixed
M mass flow rate
c specific heat
Increase M you decrease (T0-Ti).

 

[rmw]

Zekeman,

One of the few takeaways that I remember from my first university Heat Transfer course was that heat transfer is a function of T-T-T. Time, Temperature and Turbulence. Sounds simplistic but it is at the heart of your problem.

You have reduced the time function of that relationship on the oil side, but you have increased the velocity of the oil through the tubes, so you have changed (hopefully for the better) the turbulence factor. Obviously the increase in turbulence hasn't matched the reduction of the residence time of the oil in the heater. Therefore, I would go after the temperature part of the equation.

Can you do any burner tuning that will reduce the excess air and bring up the furnace temperature, or change the way the air is introduced into the furnace so as to keep it away from the furance tubes? Oil heaters depend a lot on radiant heat transfer in the furnace (tubes) so any way that you can increase that heat contribution will help you. It may be that you ARE putting 165 MMBTU into the furnace, but you are just heating a lot of air that ends up going out the stack.

Reducing the excess air flow will also decrease the time factor on the gas side, but, of course also reduce the turbulence of the gases through the tube banks.

Also, does your heater gas side need a real good cleaning? Sooted up heat transfer surfaces aren't your friend just now.

This heater seems to be designed with real close margins if it can't handle less than 10% excess fluid flow and maintain outlet temperature.

Those Production slugs aren't going to back off. It is never their problem, always yours, so you are going to have to get creative in solving the issue. They will never take the truth for an answer.

rmw

 

[chicopee]

Can you get your hands on the original specifications and performance tests before the system was accepted? This information may help you in your argument with the plant management and speed up the cleaning contract of the heat exchangers. It should be easy to verify the return oil temperature and in all likelyhood is higher than the performance test for reasons explained above by rmw.

 

[zekeman]

"One of the few takeaways that I remember from my first university Heat Transfer course was that heat transfer is a function of T-T-T. Time, Temperature and Turbulence. Sounds simplistic but it is at the heart of your problem."

Did your heat transfer professor tell you about the law of conservation of energy, to wit

Q=M*c*(To-Ti)

which does not deny anything you said but goes to the heart of the problem. Since Q is fixed , do you deny the obvious, that the product of mass flow and (T0-Ti) is fixed and therefore there is an inverse relationship between
M and delta T. Your problem is that you fail to recognize that no matter what the fouling , the transfer source medium temperature rise will accommodate that problem by rising to
make sure the energy equation holds.

 

[rmw]

Yeah, we got to that part a few days later. After I absorbed all that, then I graduated and learnt a lot more about Hx's than I ever learnt out of a text book or classroom. I even suffered through some fired thermal oil heater (system) problems.

I don't agree that TO and Ti are fixed. If the user's process Hx's are all fouled up as the OP states, then the original TO and Ti won't be as per design or as per original (especially the Ti which is the T-return from the process), so the textbook formulae might not be applicable here.

A good cleaning or tuning just might.

If the process heaters are returning oil back to the heater hotter than normal when clean, then the whole design premis of the heater is now on a different delta T and mass flow basis.

And.... as is the case in lots of problems, the Root Cause might still not be identified here. Let's keep digging.

rmw

 

[zekeman]

"don't agree that TO and Ti are fixed."

They aren't,only their differece is of interest.

 

[Sam654]

Thanks for your input guys.

rmw
You hit on several issues we are aware of. This has since grown into a very complex issue.

First, I'm able to burn at 3.5% O2. That's not the issue. In fact, burning slightly higher O2 actually helps us. I believe this is due to the lower flame heights which affect the convection inlet temp. See below.

Second, the heater was designed back in the 60's to burn #6 oil. We recently converted to natural gas and have the capability to co-fire NG with #6. With NG, we bump up against our convection inlet limit of 1775 degF. Natural gas has a lower flame temperature and therefore we don't get the heat gain in the radiant section as we do with oil. This is indicated by a decrease in furnace outlet (convection inlet) temperature when an increasing amount of oil is introduced. i.e. We get more radiant heat gain from the oil and actually use less fuel. The problem being the ungodly cost of fuel oil compared to NG pushes us to use as much NG as possible for reduced cost despite the increased efficiency of burning oil. To get the same duty from the heater, more heat gain is required in the convection section but, in fact we don't due to the constraints of the initial design.

Another effect of burning natural gas is that we are limited by state DEP regulations to not exceed 165 MMBTU/hr and in this calculation, we have to use the HHV of NG and naturally only the LLV is usable. That puts us at an ~8% disadvantage of getting our full heat input.

Third, we know we have a fouling issue with our convection section. That is indicated by a steady increase in our convection outlet temp since last July. Further investigation indicates we may be receiving oil with a higher contaminant level, particularly ash and calcium. Talking with our suppliers and other #6 oil users, this is, and will be, the case since 6 oil is residual and they "milk" more and more of the "good stuff" out in refining. Sootblowing isn't removing the buildup sufficiently. We have looked at soot blower operation and supply pressure. We also think there is a "white effect" from the calcium in the radiant section which is further inhibiting our radiant heat gain.

We are looking at an online cleaning system that uses dry ice to clean the convection section. Cleaning of the radiant section will require a shutdown which is VERY costly but may be our only recourse. Obviously, we're between a rock and a hard place.

But... back to my initial question. With all of these issues relating to both radiant and convective HT, it makes sense that an increase in flow does affect HT due to decreased residence time since we're running on the ragged edge of heater capacity barring any change in heater construction. Adding additional heat transfer surface affects our permitting and could cause us to lose our grandfather clause for burning #6 oil.

Zekeman
When fully "maxed out", we do see an decrease in heater outlet temp but, eventually, the inlet temp drops also as the cooler oil returns to make your Conserv. of Energy arguement valid. With this additional information, does my arguement about residence time make sense to you?

Thanks again guys.

 

[desertfox]

Hi Sam654

I agree with you and others that the increased flow rate of the oil means its not in contact with its heat source for the same length of time compared with the original flow. Therefore it cannot absorb as much heat as previously before it reaches the outlet, even if the rate of heat input remains constant.

desertfox

 

[rmw]

Ahhhh, Sam654, the nightmares return.

Somehow I suspected that fuel oil was involved. And the days of #6 being #6 are gone. #6 is now just the trash heap of the refinery, normally the heaviest of the heavies cut with the least valuable of the lighter runs. No more is it just the last cut as it was when your burner / furnace was designed.

And, yes, you are up against mission impossible regarding the lack of radiant heat gain to the furnace section. Your only answer is to add some baffling somehow someway to the convection section to enhance the heat transfer from the NG fuel there.

Since you are able to get your O2 down so low, your gain at slightly higher O2 might be higher velocities in the convection section as well as what you suggest. You might want to see how far you can go there and find out where the sweet spot is before you start losing ground again.

Zekeman's conservation formula is always true, but you are getting into the more work-a-day formula Q=UAdT with dT being the LMTD. More important is all the factors that go into the "U" part of the equation. Whereas radiant is normally a very minor part of any "U" consideration, in your case it is big. Your increased fluid velocity has affected the film effect factors on the liquid side, and NG vs FO (#6 or faux #6) have profound effects on the gas side.

About the only 2 things you have constant in this whole conversation is the limit on Q(in) and A. The rest is up for grabs.

I'd look up the formula Q=UAdt and look at all the factors in the denominator of the "U" factor and see which ones you can go after.

You can't do this until you shut down, but you might take a page from others who do this and add swirlers to the fluid side of the oil tube(s). Adding a swirl to the fluid would help the fluid side and add to the retention time in the "A" of the heater.

You may also want to check with some of the 'snake oil' salesmen regarding additives (and there are some that aren't snake oil) for your #6. The ones that work really do work and work well. They can add stuff that will make your ash more friable and easier to clean.

I am running thin on suggestions now.

But the nightmares are coming back faster.

rmw

 

[zekeman]

'Zekeman
When fully "maxed out", we do see an decrease in heater outlet temp but, eventually, the inlet temp drops also as the cooler oil returns to make your Conserv. of Energy arguement valid. With this additional information, does my arguement about residence time make sense to you?"

That statement, standing alone is not wrong to look at reduced residence time for the reduction of differential temperature ; but for my money, I think the energy equation trumps any other way of looking at it.

The problem I had was your original explanation that you gave your management was not clear and convincing but it is to me-- but I'm not your manager Maybe you should augment your explanation with the energy equation.

'My explanation to plant management is the residence time through the unit has decreased and since I cannot increase my heat input, my heat transfer to the fluid decreases."

 

[Sam654]

Point taken... thank you.

Oh, and the 8% penalty for burning NG should be ~4%. It's the DIFFERENCE between the LHV/HHV of NG vs oil.

 

[MrBTU]

Don't forget Q=Uo(A)(LMTD)
In your case, the Uo increases slightly with increased flow rate (the "turbulence" you referenced above), the A (area) remains constant, and the LMTD increases slightly because you have more temp difference between the thermal oil and the heating elements, since you're not heating the oil up as much. So, the increased flow rate will give you a slightly higher heat input Q, but since your mass flow is increasing to a greater extent, your To is still lower. Play with the two equations we've given you and you will be able to model the two flow cases and prove mathematically to the operators what is going on.