Towards a StBC Tank model

[StBarnabas]

Hi all
early days yet but am hoping for great things here but will probably time out as usual!). Here are some thoughts as to the first days results. There are currently 7 sensors placed at x=[10 40 70 100 130 145 160]; (cm) from the floor (Navitron Solarcyl  230L tank). - this hopefully will be increase to 11 shortly. Attached 1st days graph + some interpretation. Tank seems to maintain stratification well. ASHP turned off and I am hoping for a good solar day tomorrow [MrsB away so can keep hot water usage to a minimum] - the past two days have been c**p here in Northumberland; PV struggled to make 1.5kWh yesterday and 3 today (a good day is 13kWh at present). Any comments about the graph welcome

[Pat_]

It's a lovely picture! For me, though, I intuitively see red as hot, and pale colour as cold.

It would also be nice in some way to see the temperature v height on a graph, but I suppose this can only be done as a snapshot at a particular time. I'm tempted to get a set of sensors and try what you've done on my own cylinder.

[Ted]

Agree about the colours. I would expect blue as cold(er) and red as hot(ter).

Hope you have a photocopy of your cleaner's and nanny's passports!   

[StBarnabas]

Pat/Ted
colours are arbitrary - have use these as they mimic the color of a black body as it gets hotter. But will post a few options later going from blue to red is no problem
Regarding temperature vs height plots - will give it a go when I am happier that my sensors are working well - the thermal contact is not great for some of them....-

[Richard Owen]

have use these as they mimic the color of a black body

Sean,

You have to be careful with this stuff.

A long time ago I was using the perfect black body as a way of describing some radiation effect or other and Martin got all a twitter about Halle Berry.

Don't want to start him off again.

[sleepybubble]

have use these as they mimic the color of a black body

Sean,

You have to be careful with this stuff.

A long time ago I was using the perfect black body as a way of describing some radiation effect or other and Martin got all a twitter about Halle Berry.

Don't want to start him off again.

I believe that is called incitement, you can do time for that.

[StBarnabas]

Richard/Sleepy Yes I seem to remember some first year Physics Lecture in New York getting into trouble recently for using the BB term!
  Here is another possible colour option.
Part of the difficulty is that the temperature measurements are not great as thermal contact with the tank is not perfect and insulation to the air is also not perfect. I assume that T_measured =k*T_tank+(1-k)*T_attic. Yesterday I got the entire tank mixed to 65 degrees - a good day for solar and MrsB was away so was able to run off no water.  (The PT1000 sensors S2 and S3 for my TDC3 controller were at 64 degrees and 66 degrees respectively).
I solved for k (seven values) and then worked out the tank temperatures from T_measured and T_attic. This seems not to be perfect but probably now a bit better. Will gradually improve this as time goes on.

    

[Richard Owen]

I solved for k (seven values) and then worked out the tank temperatures from T_measured and T_attic.

Some people have too much time on their hands.

[Pat_]

Oh, very nice! Interesting to see how the heat input at the bottom of the tank affects the temperature all the way up. I'd assumed it would work like that. My only puzzle is why the cold water being drawn in is so warm already. Or is it coming from a warm source anyway. (Sorry I haven't read all the details of your system if you've posted them.) On my new system, one week old, the water coming in at the bottom stays at 17 deg until it has time to be heated.

[StBarnabas]

Richard
the solving procedure took about 5 mins - sadly does not work very well.
Pat
yes you make a very good point. My tank has an integrated header tank on top of the main tank - so the water is indeed preheated. On the next sensor upgrade it will also get instrumented. Good job 1-wire is so cheap (for temperature measurement anyway).

[StBarnabas]

Well the tank modelling is going slowly. There are now 11 sensors on the tank. The mark I sensors have micro-bore copper tubing protecting the sensors with very little contact area. The thermal contact is not great. The mark II sensors have a much greater contact area – a copper disk exactly the size of a 1p coin soldered to a couple of cm of copper micro-bore pipe to make a pocket for the sensor. These are in good thermal contact with the tank through the use of heat sink paste and the core packed with “rockwool.” (Once I am happy all the sensors are giving sensible readings I will up the insulation-but also want to measure the improvement so want to be sure I understand the tank under equilibrium conditions at least before I move on to dynamics).

 
Over the weekend we were away so the tank had a chance to get into equilibrium – 48 hours without water being drawn off and also with the air source heat pump off. The top of the tank was at 67 degrees when we got back Sunday evening. What is puzzling me is that the seven mark II sensors are not displaying the same temperature. They show a monotonic increase from the bottom to top of the tank of about 6 degrees which appears to be a real effect. Ideas please? Sorry I should post some photos etc.
Sean

[sleepybubble]

have to use a theasurus on your posts quite often.... however I wonder if the plateau on your monotonic observations coincides with the top of the coil maybe? It would certainly indicate a thermocline, with stratified loading both above and below the thermocline.

eg. this image of ocean temps.... from here http://en.wikipedia.org/wiki/Thermocline



looks very alike to this image of a monotonic function from here...   http://en.wikipedia.org/wiki/Monotonic_function



although in 20 minutes when the kids go to bed the most important thing will be a ....




actually having read some of the links that wookey has posted on stratification etc I am not actually a believer, whilst having a cigerette in the shed I think what your seeing is the effects of radiation above the coil and possible stratification below the coil. The thermocline is possible induced when the plume collapses due to cessation of heat input, and the stratified water of the base of the cylinder repels the disturbunce effect of the weakened plume. Very hard to explain what I mean as I don't know the proper terminology.

[StBarnabas]

Hi Sleepy

Thanks. Thesaurus Oh dear! - But many thanks for reading my posts. Sorry I got my BSc (Hons) in Physics in '77 and PhD in '81 (Astrophysics) so am a bit used to using jargon - The problem is you don't know it is jargon till you talk to someone in another field. I do indeed get a thermocline very easily - stratification I think by another name but normally on a bigger scale. I have seen thermoclines in the Med. The waters in the inner sound where I do part of my day job (between Rona/Raasay and the Scottish mainland seem very mixed). Sorry I drift again...

What I am trying to do here is to do a "simplified" tank model- sort of available to all. By dynamic I mean being able to tell how the tank will behave when water is being drawn off; various heat sources are applied - stratification, standing loss etc. etc. Paul_h had quite a nice web page which I don't seem to be able to find showing some of the basic principles for a first order system. In the StB tank case the tank will be salami sliced into about 20 levels and the interaction between each level - conduction/convection (I am assuming radiation will be not so important though will be fudged for standing losses)

 
The thing is here I am heating the tank from the bottom - though it is not a put on a cooker but a solar coil. I had thought after a few days with no draw off the entire tank would be at the same temperature but it seems not to be. Each sensor has an error but toss a coin seven times and get a head each time? - a bit of a trend I feel...

[KLD]

Sean, you probably got all the data during the two days? What was the starting point, and have you tried plotting a temperature distribution versus time? Maybe 48h were still not long enough.
Also, could there be any heat drains from the lower section of the tank?

Klaus

[StBarnabas]

hi Klaus
Temperature distributions against time - absolutely. I have been a bit reluctant to post these as I still think the sensors are at about 80% real tank temperature and 20% room temperature, if the PT 1000 sensors are to be believed. They look OK but will post raw data tomorrow - plots on my work computer...