The DEC — plant is continuously monitored by registration of about 70 measuring points. All the important process information like temperature, humidity, volume-flow, operation signals of the humidifiers, ventilators, desiccant and heat recovery wheel, volume-flow controller and the position of the air duct flaps are registered in a 10 s — time step. For the evaluation one minute average values are used.

The monitoring campaign gives the chance to have a detailed look on non-ideal plant behaviour and offers thereby the basis for optimisation. Furthermore the monitoring suits to the purpose of evaluating the customer satisfaction and of evaluating the energy performance. The important value for the customer satisfaction are the room conditions. The energy performance is characterised by electricity and fuel consumption, solar gains and the efficiency of the heating and cooling performance.

Figure 3 shows a sketch of the flow sheet of the desiccant cooling plant including the ducts to the room.

1.0 Accuracy of measurements

Evaluating the monitored data some systematic errors came up:

• volume-flow measurement of small volume-flows: The volume-flow-meter has to work reliably in a range of 500 to 10200 mF/h. This is almost impossible for one single volume-flow-meter. Therefore the small volume-flows values have a large error margin.

• Humidity and temperature measurement in air ducts with rotating elements and little turbulence gives only the temperature information of the measuring position. For the introduced SDEC plant the dehumidification of the desiccant wheel can be analysed only in tendencies. The absolute values of adsorbed water vapour are not representative because of the strong influence of the rotation on local temperature and humidity distribution /6/.

• As inlet temperature the ambient temperature, measured at the north side of the building, is used. Meanwhile additional temperature measurements showed that the inlet temperature at the entrance of the plant is higher than the north side temperature. The reason for this temperature difference are natural convection effects at the east facade of the building, where the inlet duct for the desiccant plant is located.

1.1 Plant operation

The yearly plant operation hours depend on the using times of the rooms. Both rooms are used as meeting rooms and therefore they are discontinuously occupied. In 2002 the total amount of operation hours amounts up to 1335 hours, in 2003 up to 1289 hours.

Table 1 shows the operation hours and real using hours of both rooms. Operation hours means the hours the plant is running and air conditioning the room. Using hours or hours of occupancy means the hours where the rooms are really used by people for meetings. In 2002 for example the "Cafeteria” was air conditioned for 681 hours, but used "only” 506 hours. This difference is caused by the start — up period of the plant, where the rooms have to reach the comfort conditions.

operation hours Cafeteria on Cafeteria occupied Sitzungssaal on Sitzungssaal occupied 01 — 12/2002 1335 681 506 901 623 01 — 12/2003 1289 517 336 962 731 table 1: comparison of operation hours and hours of occupancy, 2002/ 2003

For evaluating the difference the facility manager of the IHK SO is keeping note of the hours of occupancy.

1.2 Room conditions and user satisfaction

The room conditions can be compared against the requirements of the German standard DIN 1946 part II /13/. This standard defines temperature and humidity thresholds for indoor comfort. These requirements, given by the red frame, can be visualised in a temperature — humidity — diagram, as shown exemplary in figure 3.

figure 3: comfort area (cp. DIN 1946 part II) of the “Sitzungssaal”, 1 — 12/ 2003

The points represent 1-minute average values of return air temperature and humidity in the rooms. The return air values represent the whole room situation. Therefore they are controlled and evaluated. The different seasons of the year are illustrated by different symbols. Most of the measured return air values are inside the comfort area. For the summer and spring quarter some data points are lying outside. These data points are mainly exceeding the limit of 11,5 g/ kg absolute humidity but keeping the limit of 65 % relative humidity. This situation is caused by the control design, which was controlling only a threshold of relative humidity. This control design was changed in July 2003. Now there is included an absolute threshold limit.

Characterising the comfort conditions the hours of threshold exceedance were evaluated. Table 2 shows the using hours and exceeding hours of both rooms for 2002 and 2003 separately. The calculated ratio shows clearly the more difficult situation of the "Cafeteria”. Their glazed facade is orientated to east, south and west and therefore extremely influenced by external loads caused by irradiation.

	Cafeteria Occupied th]	limits exceeded th]	ratio	Sitzungssaal occupied th]	limits exceeded th]	ratio
01 — 12/2002	506	150	0.30	623	56	0.09
01 — 12/2003	336	87	0.26	731	95	0.13

table 2: absolute and relative hours of limit exceedance, 01/2002 — 12/ 2003

The main part of the exceeding hours are caused by humidity limit exceedance. As mentioned above there where some reasons within the control design. Analysing the plant performance, some other reasons could be found out. For example, the assumed efficiency of the heat recovery wheel was not reached. Looking for reasons, it was realised that the two rotating wheels, i. e. dehumidifer wheel and heat recovery wheel, have to rotate in opposite direction in the cooling case /6/ and in same direction in the winter case. Caused by the german climate the wheels were optimised for the winter case. The influence of changing the direction of rotation will be analysed in further investigations in 2004.

0 10 20 30 40 60 60 70 80 80 100 of operation hours figure 4: cumulative frequency curves for ambient and room temperature 2002 and 2003 Figure 4 illustrates the comparison of the years 2002 and 2003 in an other way. The cumulative frequency curves are evaluated using 5 — minute-average temperature values only for the operation times of the plant. As room temperature the exhaust air temperature of the air conditioned room is used. For the times, where both rooms are used the temperature of mixed air streams is chosen. Figure 4 illustrate the percentage of temperature ranges within the operation time.

Looking at the ambient temperature of 2002 and 2003 it can be seen, that in 2003 at 25 % of the operation time the ambient air temperature was higher than 25 °C, whereas in 2002 this was the case for 11 % of the time only. Looking at the room temperature, in 2002 2.5 % of the operation time the limit of 27 °C was exceeded. In 2003 this was in 5 % of the operation time the case.

SHAPE * MERGEFORMAT

In /14/ it was evaluated that in Freiburg the average ambient air temperature during summer1 2002 was 1.5 K higher and during summer1 2003 5.2 K higher than in the Test Reference Year, which was used for simulation. Comparing the results with the promised 0 — 2 % operation time exceeding the temperature limit of 27 °C of the simulation study, the realised conditions are very satisfying.

1.3 electricity consumption

The total electricity consumption for the whole monitoring period was about 26649 kWh. This value is measured in total for all plant consumers by a central electric power meter.

	07 — 12/2001	01 — 12/ 2002	01 — 12/ 2003	Total
electricity consumption	6652	9577	10420	26649

table 3: electricity consumption 07/ 2001 — 12/ 2003

The electricity consumption is in general far higher than estimated. Three main reasons

were found to be responsible for the high consumption /15/.

• High stand-by electricity consumption, which varies between 0,2 and 0,5 kW. The higher value is caused by security functions which are active in winter. For the whole year this means a standby electricity consumption of approx. 2600 kWh.

• Fan efficiency was overestimated. The calculations done in the phase of plant design act on the assumption of a constant overall fan efficiency of 0.6 (hydraulic/electric). This assumption is right for high volume flows; in this case the efficiency can be even higher. But the volume flow is varying according to the room demand and is often quite low. In 2002 and 2003 the inlet volume flow is at more than 60 % of the operation time lower than 4000 m3/h. At 4000 m3/h the fan efficiency is about 0.45 and at 2000 m3/h for example only 0.25.

•

inlet volume now in m’/h/10 figure 5: el. consumption before and after changing air duct pressure control (minutely values)

Constant air duct pressure control caused at low volume flows higher air duct pressure than necessary for delivering the air. The higher pressure must be generated by increasing the frequency and therefore the electricity consumption. This control was changed in summer 2003. Now the demanded air duct pressure is implemented as a function of the volume flow. Figure 5 shows the results of the changed control. The electricity consumption per inlet volume flow decreased significantly.

The reasons found for the high electricity consumption are not caused by solar components of the system. This means that high stand-by consumption and constant air duct pressure control could in the same way be also a problem for conventional ventilation systems. For both the solar desiccant system as well as the conventional air handling unit there is a potential for optimisation in general.

Discussing the energetic performance one should compare the solar DEC plant with a reference system. The reference system must be a ventilation system serving the same rooms with the same comfort, therefore a compression chiller is needed. Calculating the reference system return air humidification and heat recovery is used to minimise the refrigerating capacity of the chiller. The COP of the chiller is assumed to be 4. For the reference system a nominal pressure drop of 727 Pa and for the solar DEC system a value of 1382 Pa was assumed. The higher nominal pressure drop of the solar DEC system is caused by the desiccant wheel, the second humidifier and the collector field. The duct pressure drops are the same.

figure 6: comparison of electricity consumption of a solar DEC plant and a compression chiller driven reference system

For calculating the electricity consumption the above mentioned findings concerning like stand-by consumption, volume flow dependent efficiency and air duct pressure control were considered in calculation. Therefore, in figure 6, the calculated electricity consumption of the solar DEC system is similar to measured values (cooling mode, 2002).

Comparing only the electricity consumption of fans and pumps, as ca be seen in figure 6, the reference system needs less electricity. This is caused by lower pressure drops and less components. Calculating the electricity consumption for the reference system with compression chiller (COP = 4), the ambient and room conditions given in figure 1 were assumed. Part of the refrigerating capacity is covered by enthalpy recovery. The necessary heating capacity was neglected for the discussion.

As figure 6 clearly shows the electricity consumption of a reference system with compression chiller at the assumed conditions would be definitively higher. Of course it has to be taken into consideration that caused by the not optimised direction of rotation during the summer season in the last years, the assumed inlet air conditions could not be always realised with the solar DEC system.

1.4 Collector performance

The yearly air collector performance is characterised by total specific irradiation and the specific collector gains. For the evaluation of a system without any storage and irregular operation, the irradiation in times of operation is also important to know.

Therefore the efficiency, ■p, for this solar autonomous system is defined in two different ways.

П, оы = (1)

QIRR, total

n. = —Qolaa:— (2)

operation

IRR, operation

Equation (1) defines the efficiency by dividing the gained solar energy by the total irradiation. Equation (2) defines the efficiency by dividing the gained solar energy by the amount of irradiation within the times, where the plant was operating.

In table 4 these values are given for 2002 and 2003. The value of the total collector efficiency is quite low, because of the low amount of yearly plant running hours. The collector efficiency in operation times is ranging between 20 and 25 %. Taking into account, that the collector gains are only usable in the desiccant and the heating operation mode, this is an acceptable result.

	total irradiation [kWh/ m2]	Irradiation plant on [kWh/ m2]	total collector gain [kWh/ m2]	total collector efficiency	collector efficiency plant on
01 — 12/ 2002	1092	371	74	0.07	0.20
01 — 12/2003	1296	414	100.8	0.08	0.24

table 4: collector performance values

є

collector

32,4

(3)

solar ,2001—2003 P EL, coll

Interesting for primary energy aspects is the question of used ventilator electricity for "collecting” the solar gains. Therefore an electricity coefficient is defined by dividing the solar gains by the collector fan electricity consumption (equation 3). This was done for a period between July 2001 and end of April 2003. The electricity coefficient is about 32.4.

It is important to mention that the electricity consumption of the collector fan is also influenced by the pressure drop of the desiccant wheel. Its pressure drop is at least in the same range as the pressure drop of the whole collector array. Therefore it can be estimated that the electricity factor including primary energy consideration would be at least double as high if only the electricity consumption due to the solar collector pressure drop is considered.

In 2002 and 2003 nearly 19 % of the collector gains were used for heating purposes. In the heating case between 15 and 18 % of the necessary energy input could be covered by this solar input.

1.5 Cooling performance

The cooling energy is calculated by balancing the input air stream between ambient and input conditions. The driving energy input for the cooling case is solely covered by the solar air collector gains. Therefore the solar fraction for the cooling case is 100%.

The cooling process can be characterised by the "coefficient of performance”, COPthermai. This value points out the ratio of useful and invested energy.

COP _ Qcoolmg Vinlet P! hambient hinlet ) (4)

thermal (4)

solar, cooling solar, cooling

Table 5 shows the values of the cooling performance for the whole monitoring period. The thermal COP varies between 21 and 38 %. The difference between the average thermal COP value of 2002 and 2003 can be explained by different ambient conditions (cp. Figure 4) and by changed volume flow distribution of the inlet volume flow. total cooling energy plant [kWh] collector gains used for cooling [kWh] COP, thermal 01 — 12/ 2002 1263 6039 0.21 01 — 10/ 2003 3068 7996 0.38 table 5: cooling performance values

The thermal COP increases with increasing ambient temperature. This is caused by the greater temperature difference between ambient and room. Looking at the COP as function of the inlet volume flow, it was realised, that the COP increases with increasing volume flow. In 2003 the average inlet volume flow was higher than in 2002. This effect is caused by the specific configuration of the solar DEC plant. There are times where only a small inlet volume flow is needed. In case of high irradiation the energy given by the collector field is more than needed for regeneration.

A COP comparing the cooling capacity with the electricity consumption therefore required is not yet available on an annual basis. But looking at the assumed conditions of figure 1 and the results of figure 6 for a volume flow of 8000 m3/h the electricity related COP for the reference system would be about 1.5 and the one of the solar DEC system 5.2.

As already mentioned, in the summer case the heat recovery wheel and the desiccant wheel should rotate in opposite directions /6/. This will be realised within the cooling season 2004. A significant increase of the heat recovery efficiency is expected. The results will be presented in future publications.