Burst heating coil 1

[coolmedown]

I have a heating coil 138/115 Deg C flow & return temperatures, design conditions. It is only 2 weeks old and recently burst at a support strap connecting the the flow and return header.
Before the incident under low load condition the return temperature was going back at 40 deg C. Is this normal and ok?. The control valve was only approximatly 10% open during this time maintaining off coil conditions smothly.
The coil supplier doesnt think the water should be going back so low and has blamed this on the controls system.

I am trying to figure out the thermodynamics of the coil re LMTD etc...and figure out why the coil burst.
Please help!!

 

[lilliput1]

Take a close look at the break in the piping. See if it can tell you if it broke from external stress or internal. Check the wall thickness at the break. Check the OD to see if it burst from internal pressure. Check the support strap also. Did it dig into the piping? Was there excesive vibration? Was the coil pressure tested at the factory? At what pressure? Was the support adequate or was it overstressed? Is the coil protected by a freezestat? did the freezestat trip? Did the coil freeze? When the control valve close, is there shock transmitted to the coil? when the pump run, is there shock transmitted to the coil?

 

[remp]

I think having water leaving temperatures below design values is typical at low load conditions. I must have a look at LMTD again to expain.
Sounds like the coil was not strong enough to take the high temperatures.

 

[coolmedown]

Lilliput,

Do not now if it was internal or external steress. The fracture was only a pin hole. It failed where the support strap connects to the hot pipe and happened at the back of the strap. The hot pipe header is supported off the ground at low level but it was at the strap at high level where the damage was done. There is also a strap at low level but no damage here. There is no excessive vibrations and the coils were factory tested. The freeze stat was calibrated. This happend during mild low load conditions.
Can someone tell me if leaving water temperatures of 40 degC which is less than design leaving at 115 is typical. The coil supplier says NO. I tend to disagree especially when all other coils in the plant vary weel below the figure of 115 deg C.
I am trying to understand the Q=UxAxLMTD equation.???????

 

[quark]

If 138 and 115[sup]0[/sup] are coil entering and leaving temperatures at full load, then at partial load the leaving temperature should be above 115[sup]0[/sup] and not below that. There is something wrong with your controls.

Where is the control valve located? If in return, then there may be chances that coil got damaged due to high pressure.

Regards,

 

[remp]

quark,

I think your wrong there. Assume the load goes from say
15 % to 0% (no load). Then the control valve starts closing. Isnt the leaving water temperature going to approach the entering air temperature if there is no load??
Its a strange one alright.

 

[KenRad]

If you have a modulating control valve, how could you prevent the leaving water temperature from approaching the entering air temperature at low loads? As the valve shuts, the water spends more time in the coil, and cools down more. This is a normal condition. I believe that your coil supplier is trying to avoid liability in the piping failure.

A pin hole, to me, more often points to either corrosion, or a factory defect. Anyone who installs piping systems will tell you that occassionally you see a pin hole leak on new piping and fittings that can only be attributed to a problem in the manufacturing process. Usually these problems are discovered on hydro testing of a new system, but some develop later.

Two questions - What is the piping material? Is the leaking pipe in the airstream, or out of it?

---KenRad

 

[Whoops]

Whilst the thermodynamics of the coil and the LMTD equation are very interesting, I don't see that they have any relevance to why the coil burst unless, of course, it froze. If it had frozen, there'd probably be an axial split in one or more of the tubes near the return, not a pin-hole leak in one of the headers.

It was most probably a defect in the material (i.e., supplier's liability). Swarf in the water ways or inadequate chemical water treatment might have had a role (client's responsibility), but failure in two weeks sounds too fast for this. I think I'd investigate getting it brazed in-situ, which would minimize the down-time and save everyone the grief of dismantling and shipping it back to the suppliers. I'd get the supplier's faxed written agreement to this, so that it won't affect their warranty.

I suspect they'll want to minimize their costs and will probably agree; if they value your custom, they may offer to pay. If it happens again I'd expect a new coil, unless some other cause is identified. If they won't entertain it, just buy a new coil from another supplier, it will probably be cheaper than the time you'll spend in investigating and arguing.

 

[coolmedown]

Mr. Kenrad is right I think about the return water temp being lower than 115 Deg C at part load, the explaination makes perfect sence, but how do I prove that with the equation technically with the formula Q= U x A x LMTD. When on part load the U value decreases, how do I prove that the LMTD will decrease along with the return water temp also.

Mr. Whoops, I am trying to prove that having lower return water temps is normal in HVAC. The coil supplier is using the low water temp in his defence and said it went out of specifications of 115 deg C return.
Mr. quark reckons it should not go below 115 Deg C, the design return water temp.

Who is right??

 

[Whoops]

"The coil supplier is using the low water temp in his defence..."

He may be, but it seems to be a very feeble defence. It's a red herring; it is not relevant to the defect. You need to ask them to explain how a low return temperature could possibly cause a pin hole in their coil. If they can explain it, I'd really like to know their explanation.

I think that the equation you need is;
Q = m x c x dT

(m x c x dT)air = (m x c xdT) water

Q (kW) energy transferred
m mass flow rate (kg/s)
c specific heat capacity of the fluid, air or water (kJ/kg degC).

The control valves work by reducing the water flow rate through the coil, whether it's a 2-port valve or 3-port diverting valve. The 3-port valves maintain a constant mass flow rate by diverting a portion of the water flow around the coil and back into the return.

 

[remp]

Actually 'coolmedown' i think you are trying to prove you can have a low return water temp at low load conditions and that it is normal. I think the equation Q=UxA.LMTD is the key.

i dont know the relationship between the U value and the flow rate. Low flow rate decreases the U value of the coil (thermal coefficient, by how much I dont know. You probably want to know it results in a lower LMTD (log mean temp difference)

 

[coolmedown]

Whoops,

A copper strap welded to the flow and return header at the top and bottom of the coil exist for structural reasons , so I am told by the supplier.
The coil was rated at 138/115 degC only, also according to the supplier. But with current low load conditions we have return water temps at 40 degC. Should not be, according to the coil supplier, lets blame the controls company.
The hot water in the flow pipe caused it to expand or move at a higher rate than the cold return and pulled on the low temp return pipe becasue of the hugh temp difference and caused the pipe to fracture at the strap wels on the strap at high level. This is probably true but no excuse for the coil to burst. It seems the coil is not robust. Any structural elements of the coil should be flexible enough to carry varing temps. The coil supplier say it was only tested to 138/115 deg C

I need to know for sure if at low load 40 degC water is normal, even thogh it is well below the design of 115 degC
pleas help!!!

 

[MintJulep]

Whoops is right on. Ask the supplier to explain how lower than expected temperature can cause a leak?

The coil was rated at 138/115 degC only, also according to the supplier.

Any other condition will cause the coil to fail?

So when the coil was made, fresh off the assembly line, was it at those conditions? Was it at those conditions when it was transported to your site? When it sat on site waiting to be installed? Did the manufacuturer ever consider that the system might be shut down, like during the summer?

Tell your supplier that he is full of BS. You can even tell him that I said so if you think it would help.

The control system may or may not be working correctly, but that is not relavant to the failure.

 

[Whoops]

"I need to know for sure if at low load 40 degC water is normal..."

I regret I can't prove this, it's beyond my maths, which was fairly shaky when it was in regular use.

The low water return temperature does seem to be predictable from the Q = U x A x LMTD equation.

U and A are constants.

Q is the variable and so the LMTD term must be varied by the control system to give the required value of T(air off).

The LMTD term is a function of 4 temperatures, i.e., T (air on), T (air off), T (water on), T (water off).

Of these T (water on) is a constant.
T (air off) will, I think, probably have relatively minor variations.

T (air on) varies with the outside conditions and T (water off) seems to be the only term that can be controlled to compensate for the changes in T(air on).

T (air on) will go up in mild conditions, so I'd expect T(water off) to go down.

Anyone out there who's more mathematically talented & can show this with some equations?

Even if you can prove this to your own satisfaction, I expect the suppliers will argue black is white to avoid liability. If the temperature differential had caused the failure, I'd have expected a tear at the strap. I still can't see how the stress could cause a pin-hole. It was probably weld contamination, maybe they had the new guy doing the brazing that day.

 

[remp]

Whoops,

Actually the U is a variable too. It decreases with low water flow rate (low load) so there lies the difficulty in trying to prove the LMTD goes away down during low load conditions. If we knew how much the U value changes with flow it would be easier to prove then.

 

[HVAC68]

How was 40 deg C measured ? Is the measuring instrument calibrated ? Is it functioning properly ???

HVAC68

 

[SAK9]

It appears your heating coil is very much oversized.Your control valve is only 10% open which is a clear indication of an oversized coil.When a heating coil capacity is varied based on coil leaving air temperature,the return water temperature tends to approach the off coil air temperature .This becomes very pronounced in the case of oversized coils resulting in very low water return temperatures.Here is a link to a Trane article on this issue which hopefully would make things clearer.

http://www.trane.com/commercial/library/vol261/en_26-01.pdf

Such a big temperature diffrential could have caused a thermal shock to the coil resulting in failure.In my opinion only a materials engineer is qualified to pass that judgement.

 

[quark]

Remp,

You are right that I made an half baked statement and thanks for the correction. What actually I meant was a bit different but became oversmart when I was trying to oversimplify the thing.

During low load condition(i.e keeping air flow constant and coil air inlet temperature high) hot water control valve should reduce the flow keeping the temperature difference constant(or within a close range). This is what my experience has been(not necessarily for others), both for cooling and heating coils. The reason for lower than designed return hotwater temperatures may be due to mall function of valve control system(high load but valve closing).

I also see less chances of failure due to water volume changes due to temperature. For ex. specific volume of water at 115[sup]0[/sup]C is 0.00105588 m[sup]3[/sup]/kg and at 40[sup]0[/sup]C it is 0.00100789 m[sup]3[/sup]/kg. There is a difference of only 4.5%. The chances will be almost nill if we have the control valve in the outlet.

The pinhole seems to be totally a manufacturing defect and I concur with Whoops on this matter.

Regards,

 

[krb]

I agree with most of the posts, it seems like a defect in manufacturing or possibly it was damaged during installation. One question, is the coil piped correctly? Is the control valve and balance valve on the outlet side? If the on the inlet side, they can cause flow problems in the coil at low flow conditions that may have led to the leak. Something to check.

KB

 

[ChasBean1]

I just looked at this thread and what really floors me is that if this is a typical HVAC system (HW coil heating air), the temperatures are higher than I've seen on any system. I've yet to see a supply temperature of 138°C (280°F) and a return of 115°C (239°F). I have seen values closer to 138°F supply and 115°F return, however - usually around 180°F to 150°F...

I wonder if the design units are incorrect? Should these values for the coil be in Fahrenheit? Another hint that there may be excessive supply temperature is that the HW coil control valve is 10% open. If controls are set right, a 10% open command might be closed or barely cracked. If temperature is being maintained during design conditions at 10% open, the supply may be way too hot.

To look into this further, could you give additional coil info? Design airflow? Design MBH? Design water flow?