Wolfgang Streicher, Richard Heimrath
Institute of Thermal Engineering, Graz University of Technology
Inffeldgasse 25/B, A-8010 Graz
Tel: +43-316-873-7306, Fax: +43-316-873-7305,

E-Mail: streicher@iwt. tugraz. at, heimrath@iwt. tugraz. at

Introduction

Task 26 of the Implementing Agreement of Solar Heating and Cooling (SHC) of the International Energy Agency (IEA) was dealing with solar combisystems for domestic hot water and space heat demand. One of the targets of Task 26 was to compare different combisystem designs by means of annual system simulation in TRNSYS. nine different solar combisystems were analyzed in great detail using the same reference conditions for climate, heat load (space heating and domestic hot water), and target functions.

The systems were optimized using sensitivity analysis with three target functions based on fractional energy savings (see also Weiss, 2003):

• Fractional thermal energy savings (fsav, therm)

Saved conventional fuel input of the solar combisystem compared to a reference heating system (same building, gas boiler heating systems with only small DHW store).

• Extended fractional energy savings (fsav, ext)

Extension of fsav, therm by the electricity demand of the solar combisystem and the reference system.

• Fractional savings indicator (fsi)

Inclusion of penalty functions of not reaching the required domestic hot water or room temperatures.

For each system between 12 and 30 parameters were varied starting from a base case. These parameters covered climate, collector type, size, orientation, mass flows, store size, store geometries, size of heat exchangers, heights of inlets and outlets, insulation, control settings of thermostats and control strategy of the whole system. The following paper gives on overview over the general results.

The analyzed systems and the detailed results can be found in Streicher, Heimrath, 2004 and repectively in Bales, 2003, 2003a, Bony and Pittet, 2003, Cheze and Papillon, 2003, Ellehauge, 2003, Heimrath, 2003, Jaehnig, 2003, Peter, 2003, Shah, 2003.

All Reports of Task 26 are available from

• http://www. fys. uio. no/kierne/task26/downloads. html

• http://www. fys. uio. no/kierne/task26/handbook/tech reports. html

A similar study is presented in Streicher, 2003a, where also similar results were found.

Methodology

In order to summarize the general dependency of fsav, ext in solar combisystems on the various parameters analyzed in the system reports of Task 26 the following statistical approach was chosen.

The specific alteration of a parameter was defined as

(P — P )

equ. 1: APspec = ^ Pmn)

Pbase

with

APspec specific alteration of a parameter in the sensitivity analysis of one system related to the base value of this parameter Pmax maximum value of parameter in sensitivity analysis Pmin minimum value of parameter in sensitivity analysis Pbase Base value of system parameter

The base for the different parameters was chosen as follows

• mainly the value given for the base case of each system

• 90° for azimuth sensitivity

• 45° for slope sensitivity

• 1 for specific heights within storage

The specific alteration of fsav, ext between Pmax and Pmin was defines as

_ (fsav, ext, max fsav, ext, min)

spec

‘sav, ext, base

specific change of fsav, ext between Pmin and Pmax related to Pbase

fsav, ext with Pmax fsav, ext with Pmin fsav, ext with Pbase

The dependency of fsav, exton the change of the parameter is defined as

with

fsax, ext(P) dependency of fsav, ext on the change of the parameter

Using this definition, parameters with different units can be compared to each other. Of course the value chosen for the base-case and the shape of the dependency of fsav on one parameter are influencing this calculated dependency fsax, ext(P). Nevertheless the values of fsax, ext(P) give a good feeling how strong the parameter change influences fsav, ext.

In the following pictures the median of fsax, ext(P) (dependency of fsav, ext) for all combisystems that varied this parameters are shown. As this values sometimes differ strongly to each other because of different base values und variation range of the parameter, the standard variation of the values of the different systems is additionally shown.

In case that there a parameter has been only analyzed in one system the bar of the median has a light colour and the standard variation is zero. These values are not analyzed in the following, because a general dependency of fsax, ext of this parameter for different systems can not be defined.

Additionally, parameter variations, where the user demand was not fulfilled, were taken out of the analysis.

Results