A heat exchanger will be needed for most solar water heating systems, other than direct systems, to separate the fluid in the solar loop from the potable water being heated and used. Heat exchangers introduce losses in system efficiency so they must be sized and specified properly to minimize these effects.

Most heat exchangers used for solar water heating applications are the same ones used in traditional boiler systems and have a rating based on this type of heating. In a traditional boiler system 180° F water is used to heat 100° F water in the storage tank. The temperature difference between the heating fluid and the fluid being heated is known as the “approach”. In the case of boilers systems this would be around 80° F, but for solar thermal systems the approach should be much lower, closer to 10-20° F. This is because solar thermal collectors are most efficient when the water entering the collectors is as close to the storage temperature as possible. There are many types of heat exchangers (brazed plate, tube-in-shell, immersed coil, etc.) but the steps below give a generic method for how to size any heat exchanger. Once the basic aspects of the heat exchanger are sized they can be matched to manufacturers’ published data and sizing tools.

Heat exchangers should always be piped in counterflow for the best efficiency. This means that hot and cold fluids are flowing in opposing directions to one another. See Figure 1 for a diagram of a plate heat exchanger in counterflow.

Figure 1: Schematic Diagram of Plate Heat Exchanger in Counterflow

Figure 1: Schematic Diagram of Plate Heat Exchanger in Counterflow

Sizing Method

The sizing method detailed here should be used for initial estimations only. Consult heat exchanger manufacturers’ documentation to verify use with solar water heating systems, pressure drop, and other data pertinent to the design. Below is the general equation for calculating the steady-state heat transfer in a heat exchanger:

Q = U * A * LMTD

  • Q = estimated peak hourly solar collector output (BTU/hour)
  • U = heat transfer coefficient of heat exchanger (BTU/ft2 – °F – hr) – See Table 2 for common values.
  • A = heat transfer area of the heat exchanger (ft2)
  • LMTD = log mean temperature difference (°F)
    • \[LMTD = \frac{(T_{hin} – T_{cout}) – (T_{hout} – T_{cin})}{ln\frac{(T_{hin} – T_{cout})}{(T_{hout} – T_{cin})}}\]
    • Where:
      • Thin = temperature entering heat exchanger on the solar loop
      • Thout = temperature leaving heat exchanger on the solar loop
      • Tcin = temperature entering heat exchanger on the potable loop
      • Tcout = temperature leaving heat exchanger on the potable loop

        Figure 2: Graphical Representation of LMTD in Counterflow Heat Exchangers

        Figure 2: Graphical Representation of LMTD in Counterflow Heat Exchangers

Temperatures entering and leaving the heat exchanger will usually not be known but may be assumed using targeted values. In the case of counterflow heat exchangers (Thout – Tcin) is considered the “approach” which can be estimated to be 20° F. (Thin – Tcout) can be estimated to be around 30° F. If two of the four temperatures are known, or can be specified, and flow rates on both sides of the heat exchanger are known then the other two temperatures may be solved for using the following strategy:

  • Q = ṁh * cph * (Thin – Thout) = ṁc * cpc * (Thin – Tcout)
    • Q = estimated peak hourly solar collector output (BTU/hour)
    • ṁ = mass flow rate of fluid (on solar or potable side) (lbs/hour)
    • cp = specific heat of fluid (BTU/lb-°F) – See Table 1 for reference values
    • Tin = temperature entering heat exchanger (°F)
    • Tout = temperature leaving heat exchanger (°F)
Table 1: Specific Heats & Densities of Various Fluids
Fluid Typecp (BTU/lb-°F)ρ lb/ft³
20% P. GLYCOL0.96162.910
30% P. GLYCOL0.93263.560
40% P. GLYCOL0.89664.110
50% P. GLYCOL0.85464.580
60% P. GLYCOL0.80564.970
Table 2: Common U-Values for Heat Exchangers
PlateLiquid to Liquid Heating150 – 700
Tube-in-ShellLiquid to Liquid Heating25 – 200
SpiralLiquid to Liquid Heating125 – 500

Sizing Example

Size a plate heat exchanger for an indirect system using (4) SLSG-40 collectors heating a storage tank to 120° F. Assume the collectors are operating at 150° F and a 30% propylene glycol solution is used as the HTF in the solar loop.

  • Q = (4 * 40 ft2 * 1070 BTU/ft2-day)/(5 hours/day)= 34,240 BTU/hour
    • Note that an ideal 1070 BTU/ft2-day for SG series collectors is used as well as a solar heating time of a (5) hour solar-day. These are conservative estimates for the purposes of initial sizing to make sure the heat exchanger can transfer up to 100% of the energy output by the solar thermal collectors.
  • U = 300 BTU/ft2-°F-hr (estimated)
  • Thin = 150° F (estimated collector operating temperature)
  • Tcout = 120° F (desired storage temperature)
  • Thout = Thin – Q/( ṁh * cph)
    • h = 4 GPM * 60 min/hr * 0.1336 ft3/gal * 63.56 lb/ft3 = 2037.99 lb/hr
    • cph = 0.932 BTU/lb – °F
    • Thout = 150° F – [(34,240 BTU/hr) / (2037.99 lb/hr * 0.932 BTU/lb – °F)] = 132° F
  • Tcin = 132° F – 20° F = 112° F
    \[LMTD = \frac{(150 – 120) – (132 – 112)}{ln\frac{(150 – 120)}{(132 – 112)}}\]
  • A = (34,240 BTU/hour) / (300 BTU/ft2 – °F – hr * 24.66° F) = 4.358 ft2
  • Choose a heat exchanger with a minimum overall heat transfer coefficient of 300 BTU/ft2 – °F – hr and a heat transfer area of 4.358 ft2. Solar loop flow rate shall be 4 GPM, potable look flow rate shall be 8 GPM. Consult heat exchanger manufacturer data to verify heat exchanger material, pressure drop, connection size, etc. Higher overall heat transfer coefficients will require less heat transfer area, and vice versa.
  • Assuming this system were using a 240 gallon tank the tank would gain an estimated 80° F during one solar-day under these conditions (not accounting for any usage during the day). To backcheck this, divide the total estimated solar energy output by the total weight of water, or [( 4 * 40 ft2 * 1070 BTU/ft2 – day) / (240 gal. * 8.33 lbs/gal.)] = 85.63° F. The difference in the two results represents the 6.5% efficiency loss caused by the heat exchanger.