Basis of the investigated Slurries is microencapsulated paraffin, which is also already used in PCM-Construction-Materials [5, 6, 7]. The diameter of this capsules is down to some ^m and the thickness of the shell is far below 500 nm. Hexadecan with it’s melting point (Tm) of 18°C, Oktadecan (Tm = 28°C) or also Triakontan (Tm = 68°C) are examples for the paraffins, which are encapsulated. Melting points are also adjustable between the melting-temperatures of different Paraffins by mixing them [2].

The PMMA-shell prevents an interaction between the carrier-fluid and the PCM, that’s why in principle no additives are required in the slurry. So it’s relative easy to prepare such a PC-Slurry: The Capsules are delivered by BASF as 50% dispersion, at ISE the slurry is mixed by adding water to the dispersion till a pumpable consistency is reached. At the test-rig capsules concentrations up to 30% have been pumped. This concentration has still a comparatively low viscosity. The dynamic viscosity of a slurry with an amount of 25% capsules reaches from 7 to 8mPa • s1. Values of the kinematic viscosity for slurries with different percentage of microcapsules can be found in literature (cp. table 1).

Table 1: Kinematic Viscosities and density of different Slurries, Shell material: Urea- formaldehyde, from [1] Slurry Density M Kinematic Viscosity И water 997 9.07 • 10-[38] 5% suspension 983 1.04- 10-6 10% suspension 975 1.24- 10-6 15% suspension 966 1.51 • 10-6 20% suspension 957 1.90■ 10-6

A test-rig was built up to check the stability of these slurries. Fig. 1 shows on the left side a schematic of the test-rig. Three different cycles are visible: 1. The cooling-cycle (blue) with a performance of 4 kW, 2. The heating-cycle (red) with a performance of 10 kW and 3. The slurry-cycle (black) with volume-flows up to 1000 l/h and a fluid-capacity of 15 l. The heat is transferred into and out off the slurry by screwed plate heat exchangers. All temperatures are measured by Pt100 thermo-sensors, which are mounted directly on each in — and outlet of both heat exchangers. The volume-flows in each cycle are recorded by magnetic-inductive flow-meter and the pressure losses are measured by relative-pressure-sensors.

The slurries has been pumped for several weeks to verify their stability. Thus at a volume — flow of 500 l/h is adding up to approx. 33 melting — and crystallisation-cycles per hour, that’s about 800 cycles a day. After different periods samples are taken to check their quality by measuring the melting enthalpy by DSC, SEM-pictures are taken to check visually the condition of the shell. The slurry is also visually checked after finishing a test sequence.

і

Conventional components like expansion vessels, breathers and plate heat exchangers are used in the test-rig. It is also proven to run them with the mentioned PCS. Parallel to the thermal and hydraulically characterisation of the slurries it is possible to check also other components in a second test-cycle.

Figure 1: Test-rig to investigate the slurries and components

For the thermal characterisation the volume-flow and the inlet — and outlet-temperatures of each heat exchanger is measured. The primary side of heat exchangers for heating and the secondary side of the heat exhanger for cooling is operated only with pure water. With slurry on the slurry-cycle the heat-capacity (cp) of the capsules fraction is calculated with 2,4 For water the heat flow is computed as follows:

Q pri/seCwater m ‘ CPwater ‘ (Th Tc) (l)

Th and Tc is equal the the heat exchangers hot resp. cold temperature of the inlet resp. outlet. The heatflow on the slurry-side is computed for a slurry with a capsules fraction of

xcapsule to:

Qslurry xcapsule ‘ CPcapsule ■ (Th Tc) + (1 Xcapsule ) ‘ ГП • Cpwater ■ (Th Tc) (2)

Because the measured temperatures are equal to the sensible heat-flow it is necessary to recompute the latent heat by the heat transfer characteristic of the heat exchanger. To do that, first of all the heat transfer characteristic is measured with water also in the slurry — cycle. The heat flow is increased by heating up the inlet water on the primary side step by step. Than the heat flow is calculated for each temperature by equation 1. Since there are no higher heat losses with slurry at equal boundary conditions (same mass-flow and air temper­atures) the heat flow difference between the characterisation with water and measurement with the slurry (eq. 2) must be stored as latent heat:

(3)

Q latent Q slurry—cyclewater Q slurry—cycles

ДИ

capsules

(4)

Q latent m • (xcapsule )

With the adjusted mass-flow and known capsule-concentration it is possible to compute now the specific heat of fusion and to check, in comparison with the DSC-measurements, the capability of latent heat-storage into the slurry.

AT

(5)

Qsi

urry

xcapsule

Ahcapsule + CP water

area of mealting

0 ■ 0 10 20 30 40 50 60

Temperature [°C]

Slurry 50% — — Slurry 30%————————- Slurry 20% — — Capsules 100%————— Water — —

Figure 2: Enthalpy-Curves for slurries with different amount of capsules and a melting point of approx. 28° C

The capability of such a slurry to store heat is computed with the following equation:

To simulate PCS in different applications the simulation-program COLSIM was extended to compute with variable cp. For the simulation the cp-curve, which was obtained by a DSC — measurement of the pure capsules, was integrated to get the enthalpy-curve as function of the temperature. For slurries with different amounts of capsules the enthalpy-curve is computed with the following formula:

hslurry(0) — h capsuledsc (0) ■ xcapsule + hwater (0) ■ (1 — xcapsule)

Where h is the enthalpy per kg and 0 the temperature in °C. Fig. 2 shows different enthalpy curves for slurries with different amounts of microcapsules. Unlike the cp-curve, the enthalpy curve has the advantage to be a continuous increasing function thus the energy content is well-defined by temperature (cp. 2). The so called enthalpy-method is also used in other models for the simulation of PCM [3]. COLSIM has also the opportunity to compute a fluid system like collector cycles and/or heating-cycles in time-steps down to some seconds. This is the result of COLSIM’s explicit computation without iterations. Due to the small time — steps, the energy-changes from simulation-step to simulation-step are tiny what enables COLSIM to simulate with the finite difference method by inclusion of the results from the last time-step. For the energy content Q of a defined amount of a heat-carrier-fluid m at simulation time t the following simplified equation is essential:

Qt — m ■ ht-1 + Qhains — Qhosses (6)