Heat Loss in underground pipe

[Sam654]

I've been tasked with determining the savings of reducing the temperature of an underground hot water system from 230 to 180 degF at a military base. The average temperature drop on the return through the day is about 30 degF. I know the size and cumulative lengths of pipe throughout the system and the gross flow rate. What I don't know is the type, thickness and condition of the insulation. I would think the soil acts as an insulator also.

There seem to be too many unknowns. Am I correct? Is there a way to get an estimate of heat loss with the given information?

Don't worry about the side issues such as heat exchanger sizing. That will be taken care of later.

 

[MintJulep]

If you know the flow rate through a section of pipe and the temperatures entering and exiting that section...

 

[Sam654]

Thanks MJ. I wish it were that simple. The practicality of finding the pipe underground and getting the temperature and flow isn't there.

 

[IRstuff]

Why lower the temperature to below where you are with the existing thermal loss?

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Sam654]

The thinking is if we lower the temperature, the less heat loss we'll have.

As I briefly mentioned in the initial post, the heat exchangers will be too small and will either have to be replaced or some other supplemental means of boosting the temp will be required or vary the temperature up and down seasonally (operation procedures) to ensure we can provide heat and hot water to the buildings. The system is used for building heat and domestic hot water in the soldiers barracks and HQ buildings. Bottom line is "the man in charge" wants to know the numbers even if it's an estimate.

My plan at the moment is to get the supply temperature going into as many buildings as I can and come up with mean or median number then make the assumption that all buildings are at that temperature going in. I know the total system flow and the bulk supply and return temperatures. Calculate the loss and roughly double it (probably a little less) for the return side. I'll then take a ratio of current temp diffs to proposed temp diffs and see what those numbers look like.

The system demand (losses excluded) is anyone's guess at any one time. I'm trying to get a handle on the losses since that would be the savings.

I was hoping to get other estimation methods suggested before I revealed my plan to get some thinking outside of the box. Your thoughts on my plan?

 

[MintJulep]

There isn't anything "wrong" with your plan, and it seems like you have a good grasp of its limitations.

If you can get the supply for individual buildings can't you get the returns as well? That would at least reduce the number of assumptions.

Alternately, you could determine the total load of all of the buildings, and the difference between the hot made at the heating plant and the total loads is the loss.

Knowing the government I'd be a dollar that there have been energy audits on all of the buildings, and the loads already calculated. Just need to find what rusty musty file cabinet they are stored in.

 

[Sam654]

"rusty, musty file"... You can't imagine the half of it. I came from private industry and let's just say it's an eye opener how the government spends money. It's not just quantity, it's how money is budgeted and how it's delegated to be spent. Enough politics!

We do have the energy audits but again, it's who and how many are deployed but, more important is he wants these numbers soon. Virtually none of the buildings have leaving temp gauges. I know what you're thinking! These buildings were designed for their specific occupancy, not engineering or operations troubleshooting. I've asked for money to add gauges in the past and they asked "why?". In a way, I agree. There are better ways to spend my tax dollars. There are actually a few in government that care how money is spent.

Thanks for your input.

 

[IRstuff]

> For showering, very hot water is usually diluted with some amount of cold water to get to the desired temperature. If cooler water is used, then less cold water is delivered, but the loading on the hot water plumbing will go up, with attendant possibility of running out of the hot water, or not having sufficient pressure to handle the increased load. Assuming your temperature numbers are correct (I have a bit of trouble with the 230ºF water in a normal hot water supply) and assuming the typical user want's their shower water at, say, 105ºF, and cold water is at 65ºF then the first case would require 2.5 units cold water for each unit of hot. For the latter case, it's 1.9 units of cold water for each unit of hot. Assuming the same overall consumption, then the hot water delivery requirements increase by about 58%. One could possibly leave the heater at the same settings, and mix in cold water downstream.

> A brute force analysis would look at the temperature difference against the cold water, i.e., 230ºF vs., say, 65ºF, and 180ºF vs 65ºF. The ratio, then, (165-115)/165 = 0.30, so you'd save 30% on heating costs, but may have issues with loading on the water heater itself

> There will be other losses further downstream, which I neglected in the above analyses. If the water drops below 140ºF downstream, and it's used for dishwashing, there may be some loss in cleaning performance. If the water drops down to 140ºF by the time it gets to the showers, the dilution factor drops even more, which would more than double the load on the water heater.


TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Sam654]

Thanks IR. It's time I give you more of the story. The intent in lowering the hot water temperature is to lower the energy consumption at federal facilities that is mandated by the government. However, I want to make sure money saved isn't spent elsewhere to compensate for the issues developed from haphazard changes in the name of energy reduction.

We are indeed experiencing shortages of hot water in both domestic HW and dishwashing. ALL of the heat exchangers for domestic HW were sized for 320 deg water coming in. It was lowered to the current 230 deg prior to my arrival. Obviously, the new lower temp makes these heat exchangers severely undersized, which we are seeing. Tanks were placed outside of each barracks to heat the water back up during periods of low demand but, this has not fully solved the problem. The same goes for dishwashing since there is a more constant demand, especially during the day. Bad habits were developed over time that has exasperated the problem in the dining halls such as using hot water to hose down areas that the system was not originally sized to do. Those are side issues that we have to address. This savings calculation will be used to justify replacing the water heaters with larger units. If larger units are purchased or storage tanks for that matter, we have not really saved on the demand side. In my thinking, the only savings would be in pipe losses. Do you concur?

The bottom line is that the domestic demand (showers) will be relatively constant simply spread out over time from the two times that soldiers typically shower (morning and early evening) by use of storage. Dishwashing is an issue that requires a different solution other than storage volume but, those solutions will actually increase the central boiler load since the required temperature is not currently being maintained.

Just for your information, we do monitor Diff Pressure in several buildings. If DP drops, we have the ability to vary the speed of the pumps to deliver sufficient flow and pressure, usually during the heating season as one would expect. Our problems usually first manifest themselves at buildings at the ends of the three loops.

I understand my request was for a means of estimating losses. If your brute force analysis is true, what assumptions would have to be applied to make it true? In other words, what is missing in the analysis? Taking a simple ratio would give me a larger than actual savings, correct? I would use this as a caveat in my report.

As usual, your input is very much appreciated, helpful and highly regarded by me. Again, many thanks for your time and effort!

 

[IRstuff]

It may literally be a "wash;" the convective heat transfer equations are linear, while the radiant equations are dependent on T⁴, which means that the savings from radiant losses are larger than linear, so that may cover other things, like whether the water is actually colder than I assumed. If the water is, say, 35ºF, the ratio drops to ~26%



TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[MisterE2]

ouch, tried to think of how you would do it prior to simulation then ran into 2D heat transfer power law equations, you'll save yourself a lot of headaches w/ simulation + convergence criteria for heat transfer meshing. I would set the the dirt at 10m around the pipe and just run a simulation of how much heat loss you'll incur with a constant boundary condition on the soil at 10m at about 62F as its underground. If you're trying to justify simulation in the first place you could do heat transfer conduction at the dirt with a constant temperature of +/- 10 degrees of 62F depending on season, i'm fairly sure there are equations for modeling a slice of the pipe, afterwards you can show your boss the excel file of how much energy might be lost with reasonable estimates from the pipe with different amounts of insulation and say you need some time to set up the simulation to get a more accurate estimate.

Though i have to say, the process seems a bit pointless, a few degrees of water temperature diff is fairly insignificant in the energy scheme of things, you'll probably end up doing a lot of extra work since nobody seems to have calculated the cost of losing just a few degrees versus the cost of digging up the pipe, insulating it and the shut down cost.

 

[davefitz]

This is a good topic.

There are available Ultrasonic devices that can be clipped onto the pipe to measure the flowrate, and of course thermocouples can be addded, so the actual loss can be measured. Alternately, if you have the details of the soil's conductivity you can calculate the loss using a 2 D fiite element model of the piping, or use the older curves plotted in the 1930's( refer to Kreith's handbook of heat transfer).

A simpler approach is to assume the soil average temp is 55 F and prorate the heat loss as f= (180-55)/(230-55) = 71%, so you would reduce the heat loss to the soil by 30%.

A more important approach is to change the philosophy of the setpoint . Rather than use the dishwasher's 180 F requirement as the global setpoint, use the shower's 120F requirement as the global setpoint and add a booster elec heater to the dishwasher circuit- there are dishwashers ( european style) that have a booster heater built into the dishwaser.

"Nobody expects the Spanish Inquisition! "

 

[IRstuff]

Boosters are similar to tankless water heaters, and presumably, suffer similar issues.

> Electric boosters have very high operational costs, since the usage often corresponds to peak demand periods. Even when it's not, the cost can still be high compared to a natural gas heater

> Tankless water heaters in general are supposedly prone to scaling problems, requiring either periodic scale removal, or requiring softening of the water.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Sam654]

Mister E2
Thanks for the input. Unfortunately, I have to go through the exercise to prove the exercise is pointless. I've done a little digging since and have found some articles that reduce the effort down to a "plug and chug" formula but even with that, it's not all that simple. Besides the variation of the system layout (parallel loops and pipe sizes), heat load varies with the nat. log of the pipe and insulation diameters and uses a "Shape Factor" that varies with the nat. log of the depth and diameter of the pipe, so I'm assuming a linear interpolation may be well off. I have found the TOTAL amount of pipe with diameters and lengths, but no documentation on the amount and type of insulation that I have come across yet.

I dug up some pump curves that tell me the Total system flow. I know temperature going out and return... a gross measurement of system load. It matches fairly well with my boiler load which gives me some confidence in that number.

In 32 years of engineering, this is the first time I've had to do an exercise like this. My background is mainly operations and maintenance. Layoffs can force people into very different situations than they are accustomed.

IRstuff
My experience with electric heaters is very much the same as you discuss.

I'll keep you guys posted on my success (or lack of).