Outdoor reset wind sensor?

Mike Kusiak_2

I have been working for several years implementing outdoor reset control on a 1950's hot water heating system for a 100 unit garden apartment complex. So far, it has been very successful in terms of reducing fuel useage and increasing resident comfort. The last remaining challenge is compensating for the effect of wind on indoor temperatures. Currently with no wind, temperatures of 72F-74F can be easily maintained in all apartments with outdoor temps ranging from 10F to 50F.



The problem arises when the winds start to blow. The complex is located in a location subject to high winds. Depending on apartment location, some rooms can drop 5 or 6 degrees in temp, and this being a co-op with older residents, the complaints cannot be easily ignored. The current solution is to have the super bump up the reset curve a few degrees when high winds are predicted, and it works fairly well. The downside is that if he forgets, the complaints start coming, and when he forgets to turn it down again during normal weather, everyone overheats.



Is there a wind sensor available to automate this process?  I am looking for a device that would detect wind speed over a certain velocity and give me a contact closure which could be used to shift the reset curve upward. Shifting the curve is fairly easy to do since the Tekmar controllers have a setback control input which is not currently being used, that could instead be set to provide the wind temperature compensation.



The preferred solution would be to install TRV's on some or all of the convectors, but with 600 plus convectors and a limited budget, this is not feasible at the present time.



Any other possible solutions that would not require TRV's?

Ironman

Indoor Sensor?

Mike,

How about some type of indoor sensor or reset?

icesailor

Outdoor wind:

And that my friends is the biggest problem with outdoor reset running the show. While tuning a heating system so that the eagle poops pennies, along comes the laws of unintended consequences. No good dead goes unpunished.

Design a system with 15 MPH wind for infiltration and the wind is gusting to over 50 (like it is where I work at this moment) and you have a cold house. Easily corrected by an increase in water temperature but someone needs to turn up the water temperature. There must be a way.

I have had numerous occasions where a house was cold when it wasn't all that cold out. Like 28 degrees and blowing 35 MPH from the NW. The house wouldn't get over 62. The boiler was cranked up to 220, the system was designed at 180 degrees, and the boiler was cycling. 220' out and 200 coming back. A week later, it is 5 degrees, the house is at 70, the thermostat is at 70 and the boiler is at 180 and cycling. The difference? WIND. Infiltration. From places you/we have no control over. I have a nursing home I do with reset. When it blows over 35 from the NW and it is under 30, the maintenance person raises the system temperature.

Where "I" work, I meet these severe environmental conditions every day. If I did all these razor thin settings, I'd be paying out of my own pocket for service calls over cold buildings. Or, YOU do it and they call me. I know from experience what the problem is. You think I'm going to act stupid? I'm going to crank the temperature.

But all this "stuff" on a hydronic system. See how many jobs you loose to scorched air. Where they wrap the metal duct with 1/4" R-13 bubble wrap. And have one big return at the bottom of the stairs to do a two story house and a bypass to let the return balance out the system.

I don't believe in over engineering things. Keep it simple but don't be stupid. I heard that somewhere.

I could win that $1000.00 in the ugly system contest if I could send in some of the scorched air things I have seen. A true "OH MY GOD!!!" moment.

Mike Kusiak_2

Good idea Bob, but hard to implement

Indoor sense is an excellent suggestion but a bit difficult to accomplish. First of all there are three separate boilers, each of which supplies 5 or 6 detached buildings, with hundreds of feet of mains, some underground. So remote sensor wires would have to run the same way as the mains. The problem rooms are of course in end apartments, furthest from the boiler at the ends of the mains. Then, finding a single representative apartment is difficult, so averaging of 4 sensors in different buildings would have to be used. Now you are talking about literally running a thousand feet of sensor cables back to each boiler room. Gets complicated!



When I first started this project, I planned on using some form of indoor sensing. The complications involved and the fact that everything worked remarkably well without it kind of put indoor sensing on hold. The Tekmar reset controllers have the indoor sensing input, so it still can be done at any time if the details can be worked out.

Mark Eatherton

Good idea Mike...

Just a slight problem. No one makes a wind sensor that would have an output that could be interfaced into a conventional control logic. You would need your own converter that would count pulses generated by the anemometer, then make the decision as to when to close contacts, and when to open contacts. I'd also love to see someone come out with a "clear night sky" sensor, and it would essentially do the same thing. When it is determined that there are no clouds, the building is going to experience a greater heat loss, meaning the supply temps need to be boosted to compensate for the compounded heat loss. It could also be used to "ignore" calls for heat in East and South facing rooms when you know that solar gains will be pouring through the window in short order. I've talked to all of the big players in the industry, like tekmar, Honewell and Johnson Controls, to no avail. I understand that Carrier has something, but it cost like $2k or something rediculous like that.



I agree with Bob, and indoor feed back would accomplish the same thing without the need for an anemometer.



Maybe we should invent our own logic interface....



ME

icesailor

Outdoor Reset

The Tekmar sensors are wired with low voltage thermostat wire from the controller to the sensors.

icesailor

Wind sensors:

Mark, "clear nights" with no wind aren't the problem. Thereis no additional infiltration. In fact, as far as the conrols are concerned, it is being "fooled" into thinking it is colder than it really is. Thus, overheating the building. Cloudless, windless nights, you're talking about radiational cooling. No wind. If there's wind, you don't get radiational cooling.

To "fix" this on a reset curve, you need a way to make the temperature curve "jump up" for a set period while the wind/infiltration is a factor. One way it seems to me is to put a bypass sensor in a location in a building that is more suseptible to infiltration than other places. Like in an attic somewhere. Allow the infiltrated air to get to it. Normally, it would be open. When it gets cold, it closes. If you understand what I am saying. Those engineer types can't see a forrest through the trees.

Mike Kusiak_2

Maybe there is something?

Just found this by a Google search:



http://www.inspeed.com/anemometers/WindSwitch.asp



Gives you a relay closure at a selectable wind speed, pretty inexpensive, might be worth a try?



Maybe I am trying to over-engineer this and I should just raise the water temp, but that is what I was originally trying to correct. People just opening windows to regulate the temperature.

Mark Eatherton

Binga....

Good find Mike. Just what the doctor ordered. As for over thinking it, if it is negatively affecting human comfort, and if left to a human/manual option, it will waste energy, then yes, it needs to be addressed. I think you are headed down the right path.



It's just that the path hasn't been defined nor traveled yet. :-) I believe that makes you a pioneer...



ME

Ironman

Interesting...

But that thing averages wind speed every 5 seconds.  I would think every 5 hours would be closer to what you want.

icesailor

Wind speed sensor:

5 seconds check is OK, it all depends on what you set the wind speed at. If you set it at say 20 MPH, and it is blowing 20, you may just pick up gusts at 20 MPH and over. But it may be blowing a steady 15. Therefore, the thing will be cycling. If you set it to 10, and it was blowing 15 to 25, the unit would be ramped up to high wind.

Air infiltration is the single highest heat loss factor in the heat loss program I use and the wind factor will raise it the most. Air leaking around a badly weatherstripped door will cool off a house faster than you realize. Your house may be warm but someone Else's may not. If you don't appreciate wind induced heat loss, you will look foolish in the eyes of others. Like cold customers. I've been playing with the wind since I was 9 years old. Someone once said to me, "You're that big guy that goes fast in light air". There's usually wind along a shore under the trees. My son is still talking about the first time he went ice sailing and went 37 MPH in 7 knots of wind.

I'll have to start going around the country with NOAA.gov/weather looking at area winds on the ground. All I know is that when I am ice sailing in the winter, it is either not blowing and we stand around the ice waiting all day for wind. Or, it is blowing the rectal orifice out of a cow and I need my cut down storm sail. And some days are just plain perfect. Or, put another way, the chimney smoke is either going straight up or sideways.

Mark Eatherton

I think you missed my points Chris...

And you are correct, that it doesn't compound conductive heat loss through a wall. It compounds heat loss through highly glazed areas. It can also be used to correct thermal overshoot associated with solar gains coming into a room by anticipating the solar gain, and not dumping heat into a slab that will also have heat pouring through a window in the near term (4 hours).



The other potential application would be for snowmelt systems being idled. If there are no clouds to act as insulation, the slab will projectile VOMIT btu's into outer space. WIth the re-radiation sensor, you could shift or ignore the slab temperature during clear sky conditions. That, plus an interface into the local weather (NOAA) would give a system like ENV the ability to anticipate incoming snow, and warm the oven up before you start throwing bread into it.



I had a hotel engineer up in Vail Colorado who determined whether he needed one boiler or two running at night based on clear skies. If it was cloudy, regardless of the OSA temps, he could get by no problem on one boiler. If it was clear sky, he HAD to fire the second boiler, or guests would start complaining. As I remember, he had two towers, about 15 stories each, all glass, all the way around.



Most custom houses up here have HUGE amounts of glass to allow the occupants to soak up the fantastic mountain views. I can counter it with my radiant windows, but I don't have my radiant windows in every situation, and retrofit of the windows is a real booger to do.



Night sky re-radiational cooling has a significant impact on comfort and energy consumption where highly glazed buildings, and poorly insulated buildings are concerned. ASHRAE has recognized this. Now if we can get the control industry to recognize it, we'll have it made.



ME

Mark Eatherton

Throw a time delay relay on it....

This way, the system doesn't keep flip flopping, causing expansion/contraction noises at every gust. Wind is wind, regardless of how fast it is moving, and it will compound the infiltration of most mortal houses.



If wind is sensed, lock it into a boosted reset program for 4 hours. If at the end of four hours, if the wind is still there, it stays locked into the boost mode. When the wind stops, it goes to the unoccupied settings. With TDRs, if they get another signal when they are already in TD mode, they just reset their timers to start over.



With tekmar controls, a person could use the occupied and unoccupied setting to achieve this. You would program the control for windless operation, and a normal ODR reset using the unoccupied setting, and then when wind is sensed, you would go to the occupied, or boosted settings. When the wind stops, everything resets automatically, instead of depending upon the double hung thermostats to maintain good human comfort.



I think this idea has wheels...



ME

heatboy

Or.............

.............. you could actually tighten the building so wind has little effect. 



Short of that, raise the slope, or DWT, and cycle off the thermostats.



If the wind makes that big of a difference, reset is not your problem.   Speaking of wind, it sounds like you are peeing into it.  :-)  

Mark Eatherton

Good idea Jeff...

Unfortunately, most of these organizations have ZERO money for conservation, but have plenty of budget for fuel expenditures. Been there, seen that. Especially in a senior coop housing project like this. I've had board members of similar situations tell me, "Why should I invest in conservation? I'll be dead and gone before the thing pays for itself. I just want to be comfortable...." Tough to argue with that.



Common sense would say otherwise, but unfortunately common sense is RARELY employed in the real world.



Belated Merry Christmas BTW.



ME

Ironman

unfortunately common sense is RARELY employed in the real world.

"...unfortunately common sense is RARELY employed in the real world."

                                                                           - Mark Eatherton





How true, how true.

heatboy

I do agree..............

........ and I know how the world works.  That said, Mike is having this issue and no doubt spending valuable, uncompensated, time to make this work, when in all reality, it won't and there is no benefit to him.   The building needs improvement to make the system work properly.  



I know if I'm going to get calls to that property when it gets windy, the slope or DWT gets bumped up and stays there until, when and if, the building needs are addressed. 

Mike Kusiak_2

Time delay relay

I like the idea of the delayed off relay, Mark. I was a bit concerned about the possible chattering with changes in wind speed. With adjustable delayed off, it would really be like the human control now, but eliminating the "forgetting factor"  Sense wind and then boost water temp for a certain period of minutes or hours.



The buildings are actually pretty tight. 1950's wood frame with one course brick veneer, new vinyl double pane windows. Indoor temps only drop 5 or 6 degrees worst case with 30-50 MPH winds and no indoor feedback. Everybody is happy with room temps in the 70's, but when it drops to 69 or 68, the calls start. No complaints at all from residents during calm weather. The ODR without indoor sensing really works remarkably well, considering.



I think what I will try first is to connect a toggle switch marked "WIND" to the setback unoccupied input, and program the Tekmar so unoccupied is normal and occupied is boosted. If this works out well with manual control, we can then proceed with the wind speed sensor.



Thanks for all the input!

Ironman

Timer instead of Toggle

Mike,



How about a simple rotary timer switch instead of a toggle. That way you won't have to worry about forgetting to turn it off.

rlaggren

Cabling problem - alternatives?

The house automation people have been sending signals through house 110v wiring for a long time. There are problems at "interfaces" like breaker boxes and service transformers but some are surmountable; the utility also has opinions(!) about what goes over it's lines but the technology is presently in use. Might be easy to just try it - plug in a transmitter in

one of the distant rooms and see if you can "hear" it at the boiler. I'm not familiar with this area though so can't go beyond pointing it out.



The other way is wifi - just like the internet. I believe that again, the home automation people are already there. Ten years ago I looked into what it would take to monitor temps in tenant's apartments and the technology was available. IIRC it takes one computer which queries each sensor periodically by calling it up over the internet - doesn't care where it's located, just what it's address is; various levels of decision logic drive serial ports controlling relay boards. Directional antennas are about $100-200 ea (many or less) and you might need one or more; you need a "bridge" router which can see both the sensor and the base or if the base is close enough to the sensor it can reach it directly. IIRC sensor-monitors are sold be several companies, price depending on how hostile the environment will be and misc features - but are in the $100-200 range, maybe less now.



Either of the above might be subbed out. As a semi educated WAG, the base with the first connection might be $2000 (all told, turn-key) with additional at maybe $300-500 ea. Maybe. I definitely don't know. But the technology is out there on the shelf and costs _should_ be somewhere in the range where multi-building establishments could consider them seriously.



Rufus

CC.Rob

the envelope, please

First, do whatever you can for the building envelope to reduce infiltration. For what some of the tech approaches here might cost, you could do a whole lot with a canister of two-part foam and some caulk. Recently worked with some seniors in a drafty house. One tube of caulk solved a whole bunch of problems in a day room.



Second, figure out a way to get bulk indoor sensing. You can ladder indoor sensors with tekmar gear to get an average. The air sealing will help drive all the rooms toward that average. Do you have a spare pair of wires in the thermostat runs? Also like the idea of a wireless solution. Maybe ask tekmar what they suggest.



Indoor sensing integrates all the various losses and gains. The indoor temperature and its trends define the load. So when you have a party on a cloudless night, your heating system doesn't overheat due to the 20 people (at 200-250 BTU/hr each) in your house.

rlaggren

"ladder indoor sensors" means...

Sensors are wired together locally within a certain limited area and you get one output from that whole group?



Rufus

CC.Rob

here tis

http://www.tekmarcontrols.com/sb/sb018.pdf

Mike Kusiak_2

Sensors

Yes, laddering the sensors involves connecting the thermistor sensors in a series-parallel arrangement so that the array has the same equivalent resistance as a single sensor. The simplest array is a set of four in series-parallel. This would be a good solution if I could easily get the wiring run. As I mentioned previously, each boiler feeds 5 or 6 separate detached buildings, with a total of 30 to 40 apartments, so I would have to select a "representative" apartment in 4 different buildings. The next possible sensor array would be 16, which would give very good averaging, but would further complicate the wiring.



Here is a wireless system I have been looking at which might offer some advantages for temperature monitoring.



DataNab Solutions: BarioNet Wireless Gateway for Wireless Temperature & Humidity Monitoring