Horizontal Tank - Performance Data

Some may find this data interesting. This is a follow-up to Horizontal Tank - Stratification posted a little over a month ago. Look at that post for more information on the tank and sensors.

After a full tank charge, over a 3 day period I logged tank standby heat/btu loss and tank heat/btu's used to heat the radiant floor in my shop. Most standby losses are also to the shop. The tank has insulation consists of 6" fiberglass (R19) + 2" foam (R10) + 6" fiberglass (R19) on one exterior side wall and one exterior end wall, and all other tank box surfaces are interior and consist of 6" of fiberglass + 2" foam insulation.

Floor temperature (FL) ranged between 64-67F. Inside temperature (IS) ranged between 61-69F. Outside temperature (OS) ranged between 25-46F.

A few observations. 1) The boiler firing added considerable btu's into the heated space. During the charge period, inside temperature rose from 65-69F while outside temperature was falling from 44-37F and no heat was being drawn into the radiant floor. This data argues well for installation of a boiler system in the heated space rather than an out building.

2) Standby 1 was for 22.5 hours, and the only heat to the shop during this period was from standby losses of 4,819 btuh (less btu's lost to the outside walls).

3) The Draw btu's shown on the chart are only for the draw period. If Draw 1 and Standby 2 are combined, a period of 9 hours, then total heat load was 16,034 btuh. Similarly, for Draw 2 and Standby 3, total heat load was 17,514 btuh. Similarly for Draw 3 and Standby 4, total heat load was 9,340 btuh.

4) During the log period Standby 1 - Standby 4, heat load was 9,814 btuh at outside temperature of about 35F average and inside temperature of about 65F (deltaT= 30F). After the shop was built, I calculated heat load at 40,000 btuh (outside temp of -30F and inside temp of 65F. Logging data again during a cold winter period will provide the data to know what the heat load actually is.

5) Standby losses are proportional to deltaT tank temperature and the ambient surrounding temperature; the higher the deltaT, the higher the standby losses, and vice versa, which would be expected. But nearly all standby losses in my installation are into the heated space, the exception being losses through the nominal R48 outside walls and to some extent into the floor.

6) Standby losses at roughly 2,500 to 5,000 btuh are significant and also argue well for installation of a storage tank in the heated space.

7) The benefit of the 1000 gallons of storage is pretty clear. The total period covered by the chart was about 2 days, 18 hours, and only one boiler firing of about 5 hours with no boiler idling. The wood burned was all pine slab wood.

Here is the chart I forgot to post.

Good insight! That cool pool at the bottom is not much water, either. The bottom 12" of the tank totals about 100 gallons, the "bottom" sensor is on the bottom metal, and the sensor at the 12" level (L3/R3) reads right up along with all the other hot water sensors above (L2/R2 and L1/R1). The L3/R3 hot water must extend some distance into that bottom 12", so that cool pool is certainly less than 100 gallons.

One other observation - you have a VERY low temperature drop through your heat load. The return water is only a few degrees cooler than the supply. Ideally, return water would be as cold as possible.... I don’t know the details of your situation, but it would be interesting to reduce the flow rates and see if you get better storage performance.

Click to expand...

This is the situation. Hot water supply drawn from the top of the tank flows through Side A of a plate HX and then returns via a diptube to the bottom of the tank. Side B of the plate HX feeds the radiant floor (anti-freeze in this loop). Both Sides A and B have 007 circulators. Flow through the radiant loop is estimated at about 2.5 gpm (6 loops, 0.4 +/- gpm/loop). The radiant loop supply mixes down to 95-100F. An embedded sensor in the concrete floor using a Johnson A419 is set at 65F with a 1F differential (off at 66F). Radiant supply/return deltaT is about 33F. Radiant floor supplies about 40,000 btuh (33 x 2.5 x 500).

I haven't calculated the head on Side A of the plate HX, but it should be quite low: a 16' loop of 1" copper, plus the HX, plus a water filter. If for example purposes estimate at 6+ gpm, then Side A will see about a 13F +/- temperature drop from supply to return. If 170F water in, 157 water out and returned to bottom of tank. I have not yet put sensors on the HX, but that will happen soon.

As the chart shows, draw from the tank results in thorough mixing. This normally might not be good for purposes of firing the boiler and recharging the tank, as it would result in higher temperature return water to the boiler than if the tank was stratified, IF recharging was done before the tank was substantially discharged. So far I have been discharging the tank down to 90-120F before firing the boiler, and in this scenario I achieve a very high deltaT on boiler supply to and return from the tank. Also so far I have shut off the radiant floor while the boiler is being fired to eliminate tank mixing caused by the radiant loop.

Yesterday, for example, the tank was down to 95F top to bottom (tank well mixed, radiant just cycled off). I fired the boiler, burning 156 lbs of kindling and pine slab wood, and at end of burn tank was at 182F. Btu's added to storage were 722,100. Net btu's/lb of wood were 4,628. Based on current outdoor temperatures (30-50F), this tank charge should provide at least 3 days of heat before I need to fire the boiler again.