Aquastat on return side only?

Thor

Is it ever reasonable to control a hot water system via an aquastat on the return side only?



Here is why I ask:



* 10MMBTU steam boiler feeding a heat exchanger via a steam valve. Nothing else is supplied by this boiler.



* steam valve feeds hot water heating system serving approximately 1,000 rads on two loops fed by (very large) circulator pumps.



* current controller is a Heat-Timer HWR controlling the steam valve. I was more than a little surprised when I discovered this -- I had expected control to be applied on the steam side, not the hot water side of the exchangers.



However, this actually works pretty well a lot of the time. Rather than picking a fixed burner duty cycle from the outdoor temperature like a steam controller (Tekmar 279 looks like a good candidate) would have to, the HWR picks a target hot water temperature, controls the steam valve to get it, and the pressuretrol cycles the burner. This seems like it should give pretty good control without a lot of manual fiddling.



And it does -- except when the wind kicks up. The building is in a very high wind location, is quite old, and suffers huge infiltration -- hundreds of thousands of BTU difference at the same temperature depending on wind speed. This makes it impossible to select a heating curve that's aggressive enough under high wind conditions without significantly overheating when the wind dies down.



I know there are many standard solutions to this problem. All are difficult to implement here for various reasons. I am wondering about an unusual one.



What if we moved the aquastat from the supply to the return side, and either selected a different curve or offset the current curve to match up to the expected return rather than the expected supply temperatures? This seems like it would "automatically" account for the differences in heat loss due to conditions other than outdoor temperature, while still using outdoor temp to pick the set point.



There must be some very basic reason why it's not done this way. Please, educate me.

Charlie from wmass

There are controls that take reading from both

these would work even better as they can ramp up or down the heat depending on the delta T of the system water. Variable speed pumps could also be looked into. this could save electricity as well as fuel. How large are the current pumps and do you have any idea what they are required to pump for your system to work on a design day?

Thor Simon

Variable speed pumps

One piece of data I don't have is the pump sizes. They have somehow lost their nameplates. It's a good think I went looking, in the process I discovered a pile of other inaccurate information in the "engineering survey" of our heating system we had done a couple of years ago (I was curious why the pumps were not mentioned).



How would control arrangements for variable speed pumps generally work? Retain the current boiler controller with something else controlling the pump speeds based on temperature delta, or...?



I've certainly seen controllers for smaller hot water systems that read both supply and return temp. Looking at the way the current Heat-Timer controller is hooked up, I don't really see anything comparable that offers this feature. Will any hot-water controller work, within reason, with appropriate wiring to run a steam valve instead of controlling a burner? The HWR/HWRQ are rated for this use per their datasheets; looking around briefly I haven't found anything else that seems to be.

Charlie from wmass

You are in the big leagues with your set up.

I am guessing in the city limits. Why not get one of the pros Dan has listed to come through and take a look. They will be more familiar with your options in your area. Heat timer has some newer options. Your current system is treating the steam valve like a burner circuit and just adding heat that way. It really does not care the heat source. The Steam side may need some tuning too.

Thor

The pros!

We intend to. But we have had very bad luck when we've obtained advice in the past without having some understanding of the basic parameters of problems first.



I'm glad -- overjoyed -- to get the right advice from the right people. But I need to have some understanding of where they're coming from, first. I just almost wasted quite a bit of one of those pros' time sending them a professional engineer's detailed report on the current state of our system -- which turned out to be full of details that were just plain factually incorrect.



When I did what the engineer clearly did not do, and went to the boiler room with camera and sketch pad in hand to make some pictures, the question "hey, what if the sensor were on the return, not the supply" bubbled up to the top of my mind... so I asked here. :-)



Unless I am misreading the manual, by the way, even the newer Heat-Timer Platinum HWRQ does not use a return temp sensor. It can read ambient temperatures, of course, and shift its curve accordingly; there are reasons why that might be very tough to implement here but it is something we will, indeed, ask a qualified professional about.



Thanks!