4.3 Influence of Pipe Connections at the Tank on the Energy
Fig. 4–4 Reference test: measured.
0 10 20 30 40 50 60
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Time [hours]
Temperature [°C]
T7 T6 T5 T4 T3 T2 T1
31.6
19.6 15.3 51.5
Fig. 4–5 Reference test: calculated.
Fig. 4–6 Circulation test: measured.
0 10 20 30 40 50 60
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Time [hours]
Temperature [°C]
T7 T6 T5 T4 T3 T2 T1
39.0
30.9 27.4 25.3 51.2
Fig. 4–7 Circulation test: calculated.
For both tests seven temperatures inside the tank were measured and plotted. Now with TRNSYS the experimental setup of these tests was modeled (Fig. 4–5 and Fig.
4–7) and with this tool it is possible to model the PEX pipe like an internal heat exchanger. The goal was to find the correct UA-value for the PEX pipe. Based on theoretical calculations using the theory of heat transfer for free convection (based on the dimensionless numbers of Nusselt, Grashof and Prandtl) for the 2016 PEX pipe
with a length of 1.45 m, 50°C inside the pipe and depending on the tank temperature UA values between 10.2 (tank temperature = 40°C) and 11.3 W/K (tank temperature
= 20°C) were found.
With parameter fitting based on several TRNSYS simulations it was found that the UA value of 10.5 W/K for the 2016 PEX pipe fits best to get the most similar temperature curves compared to the experimental results (Fig. 4–6 and Fig. 4–7). This fitted value also is in the range of the theoretical calculated UA values.
4.3.2 Description of the Simulation Models and the Parameters
Based on these findings it can be stated that the TRNSYS model is simulating the reality with a very high degree of similarity. As the next step now the model of the solar combisystem was set up to be able to do the investigations which are in fact the goal of the whole work.
The model is in principle built up as described in chapter 4.1 (page 47), but the control concept for these investigations is different. The maximum power of the boiler is limited to 24 kW and the boiler set point temperature is 65°C all the time. Further, the auxiliary volume is always kept at 65°C as well.
As shown in Fig. 4–8 it is assumed that the system is situated in the house in a technical room, most likely in the basement. Practical experience is, that because of a lot of specific possible reasons the installers pass the pipes from the tank to the ceiling, further to the wall and then downwards to some components like a heat exchanger, the boiler or the pump and mixing unit. Based on this circumstance some metres of piping can be summed up which in most cases cause energy losses since they are heating a room which is very effectively cooled by fresh air coming in by a hole in the wall to be used in the boiler for combustion. This of course causes remarkable energy losses which not always simply can be calculated to heat the house and therefore not to be real energy losses. Due to the different designs of the pipe connections, different lengths of piping have to be calculated.
Solar primary loop
Boiler
Solar secondary loop
Boiler loop
Load loop:
DHW + SH Solar Forward Solar Return Load Return Load Forward BoilerReturn BoilerForward
Boiler Solar
primary loop
Solar Forward Solar Return Load Return
Solar secondary loop
Load loop:
DHW + SH Boiler loop Load
Forward
Boiler Return Boiler Forward
Fig. 4–8 Solar combisystem with pipe connections at the bottom (left) and pipe connections at the tank side (right).
In Fig. 4–8 on the left the design with all pipe connections at the bottom and the use of only internal pipes is shown, therefore this system concept in the following has the abbreviation “int”. These internal pipes are PEX pipes, which is a widely used plastic material for pipes inside water tanks. In Fig. 4–8 on the right the design with the pipe
connections for the load forward pipe and both boiler pipes external on the side of the tank is shown, therefore this system concept in the following has the abbreviation
“ext”. In the simulation model it is assumed that all the pipes inside the insulation of the tank have the same thermal behaviour as they would have, if they were mounted outside of the insulation and being insulated like all the other pipes. But no thermal bridges are taken into account for this kind of pipe connection on the side of the tank.
This can be argued since in this particular case of tank producer the insulation has a quadratic cross section but the tank itself is round. The pipes are connected in the corners of the quadratic casing where the thickness of the insulation is much larger than at the thinnest part, which is the insulation thickness in fact used in the model for the whole tank.
In general in the model of all different variations all pipes outside the tank are copper pipes 22x1 with 20 mm insulation, which results in a heat loss coefficient of 3.8 W/Km2. The pipe length of the solar primary loop (from collector to heat exchanger and back) is 20 m and the solar secondary loop (from heat exchanger to the tank and back) is 11 m.
The solar forward pipe and the load return pipe inside the tank are assumed to be stratification pipes. Because this configuration in all different designs does not change and also because of limitations of the tank model of the simulation program TRNSYS these two internal pipes are not modelled as an internal heat exchanger.
The length of the boiler loop is depends on the system concept, with internal pipes the length is 13 m (Fig. 4–8 left) and with external connection it is 15.5 m (Fig. 4–8 right).
The length of the load loop is also depends on the system concept, with internal pipes the length is again 13 m and with external connection it is 14.5 m.
For the concept with the internal pipes two different pipe dimensions with different heat transfer coefficients (when modelled as internal heat exchanger) were investigated. The abbreviation 2016 indicates PEX pipes with outer diameter of 20 mm and inner diameter of 16 mm. The heat transfer coefficient for this pipe was calculated to be 12 W/K. This value is somewhat higher compared to the one calculated in chapter 4.3.1 because the pipe is slightly longer (1.6 m instead of 1.45 m) and the temperature inside the pipe here is 70°C (only 50°C before). The abbreviation 3216 indicates PEX pipes with outer diameter of 32 mm and inner diameter of 16 mm. The heat transfer coefficient for this pipe was calculated to 4 W/K.
To investigate the difference of the two system concepts also for different system sizes, three different collector sizes and tanks were defined.
All three tanks had the same height of 1.6 m and also all connection heights were constant. This fact maybe might not be perfectly realistic, but it ensures to eliminate height effects on the behaviour of the internal pipes acting as internal heat exchangers.
Based on this decision the three different tanks have total volumes of 300, 500 and 1,000 liter, the auxiliary volumes are 90, 150 and 300 liter respectively. The tanks are insulated with different insulation thickness, depending on their size. The 300 liter tank was calculated with 50 mm, the 500 liter tank with 80 mm and the 1,000 liter tank with 100 mm insulation respectively. Therefore, the overall heat loss coefficients of the three tanks are for the 300 liter tank 2.70 W/K, for the 500 liter tank 2.73 W/K and for the 1,000 liter tank 3.79 W/K respectively.
The collector size was kept in a constant ratio to the tank volume of 50 liter per m2 collector area. This results in 6, 10 and 20 m2, respectively.
4.3.3 Simulation Results
Using the above described models a lot of annual calculations with different parameter settings were done to investigate the influence of the different concepts how to connect the pipes to the tank. The following parameters were changed:
Size of collector area and tank volume:
1) 6 m2 and 300 liter 2) 10 m2 and 500 liter 3) 20 m2 and 1000 liter
For each size again the type of pipe connection was changed:
1) External pipe connection 2) Internal pipe connection
Last but not least for the internal pipe connection again two different pipe dimensions were simulated:
1) PEX pipe 2016 (da = 20 mm, di = 16 mm) 2) PEX pipe 3216 (da = 32 mm, di = 16 mm)
So all in all 3 x 3 = 9 simulations (plus some special investigations) were done.
The system with external pipe connections, without collector and with the 300 liter tank, with only the top 90 liter heated, was simulated to be used as the reference. This system is very close to a typical heating system used in Danish houses.