Presented at Solar Energy Society of Canada Conference August 1999, Edmonton, Canada
QUEEN’S UNIVERSITY SOLAR CALORIMETER – DESIGN, CALIBRATION, AND OPERATING PROCEDURE
Stephen J. Harrison and Michael R. Collins Queen’s University, Department of Mechanical Engineering Kingston, Ontario, Canada, K7L 3N6 E-mail:
[email protected] or
[email protected] Tel: (613) 533-2591, Fax: (613) 533-6489
Abstract This paper describes the design, calibration, and use of a new solar calorimeter for the Canadian environment, which has recently been built at the Solar Calorimetry Laboratory, located at Queen’s University, Kingston, Ontario. Design of the calorimeter was aided by previous experience in calorimeter construction. As such, the calorimeter has many features that have proven successful in other calorimeter designs, along with some innovations. This experience also suggested that a small calorimeter, that incorporates several key features, could enable rapid testing of small window samples. The calorimeter was designed to be portable, and has been used both in and out of doors. This facility should allow for the accurate determination of the optical and thermal performance of fenestration, and has research and commercial applications.
1.0 INTRODUCTION The determination of the solar and thermal performance of fenestration is required for the evaluation of fenestration energy performance, estimating building loads, and assessing occupant comfort levels. Presently, there exist several methods for determining the thermal transmission (U-value) and Solar Heat Gain (SHG) of window systems. These methods are commonly grouped under calculation or experimental methods. Calculation methods include either handbook look-up tables or computer simulation models. Experimental methods are based on calorimetric measurements made under laboratory simulated environmental conditions or outdoors under real environmental conditions. Current efforts are directed towards the development of these methods in support of standards activities for rating window performance. Consequently, there exists a need for commercial test facilities able to support these activities as well as commercial testing requirements [1].
2.0 DESCRIPTION OF FACILITIES In response to the aforementioned need, a solar calorimeter was designed and built at the Solar Calorimetry Laboratory (SCL), at Queen’s University, Kingston, Ontario, Canada (Fig. 1). The lessons learned from the
construction of previous calorimeters [1] suggested that smaller, more compact designs are necessary to provide rapid testing of small samples, both indoors and out. Queen’s Solar Calorimeter was therefore designed to be small (0.646 m3), and contains features intended to increase response time. Along with its solar tracker, the calorimeter is located on the roof (4th floor) of the Mechanical Engineering building at Queen’s University (44.14o lat., 76.49o long.), and has an unobstructed southern view of Lake Ontario. This location is also favorable in that local weather patterns produce an abundance of clear sky test days and moderate wind speeds. The calorimeter has also been used indoors to determine the Inward-Flowing Fraction (IFF) of absorbed solar energy [2] for venetian blinds. Mask Wall Solar Absorber Panel
Active Thermal Guard Insulation
Test Specimen Air Flow
Liquid to Air Heat Exchanger Circulating Fans Supply Water Return Water Baffle
Fig. 1: Queen’s Solar Calorimeter. Photo of the calorimeter and cross-sectional schematic. 2.1 Mask Wall To measure net heat gain through a glazing system, a test window must first be mounted in the mask wall. This wall covers the calorimeter aperture and serves as the interface between the interior and exterior environment (Figs. 1-3).
Fig. 2: Calorimeter construction details. (a) Complete Calorimeter with mask wall, (b) active thermal guard, (c) flow loop, and (d) absorber panel. The mask wall is constructed of 6.4 mm plywood on the outer surface, 7.62 cm of polyisocyanurate insulation, and masonite on the interior surface, with a total R-factor of 3.876 K⋅m2⋅W-1. A maximum opening of 1.27 x 0.93 m is achievable while smaller windows can be mounted using modular wall inserts of the same construction as the rest of the mask. The mask has been painted flat white in order to reflect radiant loading (ε ≈ 0.30). An inner weather seal, composed of rubber, prevents significant air leakage into, or out of the calorimeter, while a second outer seal, composed of a compressible foam gasket, provides improved thermal resistance in addition to extra leakage protection. The integrity of the seal is ensured using 10 bolt points around the mask perimeter.
To determine losses, thermocouples are used to measure the temperature difference across sections of the mask wall. Each thermocouple is attached using white tape of approximately the same emissivity of the mask wall. Combining these measurements with its thermal resistance and surface area, Amask, allow the losses to be calculated as −1 (1) Qmask = Amask ⋅ ΔTmask ⋅ Rmask 7 mm Plywood 80 mm Polyisocyanurate
Mask
5 x 30 mm Butyl Rubber
2 mm Masonite
5 x 25 mm Neoprene
7 x 105 mm Plywood
15 x 82 mm Pine
Wall Thermocouple
6 x 51 mm Al Plate
Thermopile
EXTERIOR
INTERIOR 6 mm Al Plate
25.4 mm Polyisocyanurate Insulation 3.6 mm Active Guard Heater
13 mm Polyisocyanurate Insulation 6 mm Al Plate
Fig. 3: Mask/calorimeter seal cross section. 2.2 Calorimeter Walls The wall construction of the solar calorimeter is designed to reduce heat loss through the use of an active thermal guard (Figs. 2-3). Ideally, by eliminating the temperature gradient across the wall (ΔTwalls ≈ 0), it should be possible to eliminate heat flux (Qwalls ≈ 0). In addition, an active thermal guard reduces the thermal capacitance of the system, thereby reducing calorimeter response time. A series of five individually controlled 125 W⋅m-2 heaters, nine dedicated type T thermocouples, and 9 differential thermopiles, each with 2, 3, or 4 junctions, make up the active thermal guard. The heaters are activated when a temperature gradient is measured between the interior surface of the calorimeter and the heater. The data acquisition system will turn on a section heater if the average temperature difference in that section is greater than 0.2 oC. Likewise, the heater will turn off below 0.19 oC. This offset exists to reduce the chance of inadvertently heating the calorimeter cell due to overshoot in the guard heaters. The interior surface construction consists of fiberglass over a layer of 2.54 cm polyisocyanurate insulation. This has been placed over an aluminum plate, the guard heater, 1.27 cm polyisocyanurate insulation, and finally, an outer aluminum plate. The aluminum plate located behind the heater serves the dual purpose of providing a rigid mounting surface for the heater and ensuring even distribution of heat. The R-value of the wall from the interior surface to the heater plate is approximately 1.27 m2⋅K⋅W-1, and the total wall R-value is approximately 1.94 m2⋅K⋅W-1. The interior surface has been painted black for high emissivity (ε ≈ 0.90). The active thermal guard produces a small, but oscillating, temperature gradient across the wall. To estimate losses, the average temperature gradient over the course of the test was used. It should be noted that the R-factor in this case is from the calorimeter interior to the guard heater. The total heat loss from the entire calorimeter wall can be estimated as −1 (2) Qwalls = ∑ Awalls ⋅ ΔTwalls ⋅ Rwalls
It should be noted that there are some limitations associated with the active thermal guard and the calculation of losses through it. 1) In cases where the internal temperature drops below the external ambient temperature, the heaters will not work, and heat will be gained from the environment. Thermopile readings combined with the wall R-values, however, makes heat gain easily quantifiable. 2) Overshoot in temperature and response lag will cause some inaccuracy. Based on the performance of the wall heaters thus far, however, total wall losses appear to be below 2 W, or below 0.5 % of total metered energy. 2.3 Flow loop Heat extraction (or addition), and interior temperature control is primarily accomplished through the calorimeter flow loop (Figs. 1-2). Within the calorimeter, conditioned fluid is added to the internal circulating loop. This loop consists of an air-to-fluid heat exchanger, and the solar absorber plate. Circulation is provided via a circulation pump. The absorber plate is the primary energy absorption device within the calorimeter. Essentially, it is a plate heat exchanger placed within the calorimeter to intercept solar radiation. It is constructed of vertical copper fins connected to copper pipes (which are part of the flow loop). It is painted matte black to increase absorptivity, and is oriented to intercept and absorb all solar radiation incident in the chamber. The temperature of the fluid in the absorber serves to regulate the interior temperature of the calorimeter. An air-to-fluid heat exchanger aids in removing heat energy from the air, promotes a uniform air temperature within the test cell, and increases the response time of the calorimeter. Fans draw air from the bottom of the calorimeter, then force it through the air to liquid heat exchanger, and down across the specimen. An internal baffle ensures the direction of airflow within the calorimeter. Four shielded thermocouples meter the uniformity of air temperature within the test cell. To determine the energy removed (or added) by the internal flow loop, a reference heat source or calorimetric ratio method [3] is used. This method utilizes an electrical heater installed in series with the calorimeter loop. Recognizing that the same mass flow rate exists through the reference heater and the calorimeter, then −1 −1 (3) Q flow ⋅ (C ρ ⋅ ΔT flow ) = Pref ⋅ (C ρ ⋅ ΔTref ) where Qflow is the energy removed from the calorimeter through the flow loop, Pref is the power input to the reference heater, Cρ is the specific heat of the circulating fluid, and ΔTflow and ΔTref are the temperature rises across the calorimeter and reference heaters respectively. Finally, if we assume that the variation in Cρ is small, Qflow can be determined by (4) Q flow = Pref ⋅ ΔT flow ⋅ ΔTref−1 2.4 Solar Tracker The Solar Calorimetry Laboratory maintains a computer controlled sun tracking test frame. The device has two tracking axes to track the azimuth and altitude of the sun. The solar tracker is controlled using a VisualBASIC based program, which calculates the suns position in the sky, and controls the tracker position using two potentiometers. The tracker software provides excellent and versatile control of the tracker. It is able to fix the solar position to an accuracy of ± 1°, and has the ability to track azimuth only, altitude only, or track at an offset. 2.5 Other Systems A number of systems provide important weather data for solar heat gain testing. Next to the calorimeter, a weather tower contains instrumentation for measuring wind speed and direction, ambient temperature, and relative humidity. In addition, two Eppley pyranometers are attached directly to the mask wall for diffuse and direct radiation measurement. A pyrgeometer can also be attached if long wave measurements are desired. Control of the active thermal guard and data acquisition for the calorimeter is provided by a Sciemetrics model 641 and model 7000 data acquisition system monitored using a PC computer running a QMON [4] program. This system is simple to use, and easily customized to the user and test requirements.
3.0 CALIBRATION Calibration was performed on all metered components of the calorimeter and the system. Thermocouples and thermopiles were all analyzed. Voltage inputs into the DA system was also calibrated by the manufacturer. Full systems calibration has been performed under laboratory conditions. Using the mask wall with no installed specimen, heat lamps were installed within the calorimeter, and the power to these lamps was monitored. Tests were then performed to determine calorimeter accuracy, and response time. • Accuracy: Calorimeter accuracy was determined by comparing steady state output readings with metered power input. Tests were performed over a range of inputs from 50 to 600 W. Results indicated that the calculated power remained within 2% error of input power levels. • Response Time: The time constant, τ, of the calorimeter can be determined by observing the response of the calorimeter to a step input of power. If this response is assumed to be 1st order, then τ is the time taken for the calorimeter to read 63.2% of full output. Using both heat up and cool down step responses, at two different power inputs, the calorimeter time constant was determined to be 7.4 minutes.
4.0 TEST PROCEDURE The instantaneous energy flow rate through a glazing system, Qinput, is calculated as the difference between the gain due to solar radiation, F⋅G, and the heat loss due to the interior/exterior temperature difference (5) Qinput = F ⋅ G ⋅ A f − U f ⋅ ΔT f ⋅ A f where F is the Solar Heat Gain Coefficient, G is the irradiation level, and Af is the area of the specimen. Uf represents the windows overall heat transfer coefficient, and ΔTf is the temperature difference across the window To calculate the energy input into a calorimeter due to energy flow through a glazing system, careful metering of the input and output energy flows is required. This includes energy removed by the flow loop, energy added by any internal fans and pumps, and losses through the calorimeter walls. Energy input, Qinput, is then calorimetrically determined by (6) Qinput = Q flow − Q fan − Q pump + Qwalls + Qmask where Qflow, Qfan, Qpump, Qwalls, and Qmask denote: the energy removed by the calorimeter flow loop; the electrical power supplied to the calorimeter’s internal fan and pump; and heat lost through the walls and mask, respectively. The energy balance of the calorimeter can be seen in Fig. 4. The efficiency of a glazing system can be described as the ratio of instantaneous gain to incident solar radiation. (7) η = Qinput ⋅ (A f ⋅ G )−1 It has been shown that the time averaged thermal efficiency, η, can be graphically represented in the same manner as the instantaneous efficiency curve [5]. Therefore, for a series of tests, a plot of thermal efficiency verses ΔT/G can be developed. By using a linear regression on these points, the window system can be characterized: the slope represents the systems U-factor, and the y-axis intercept is the solar heat gain coefficient.
5.0 CONCLUSIONS Queen’s Solar Calorimeter has fulfilled the expectations of a quick, accurate, and portable test facility, able to test fenestration U-factor, SHGC, and even inward-flowing fraction. It has been employed for contract work and to support graduate student research, both indoors and out, on a year-round basis. Queen’s Solar Calorimeter has therefore proved to be a successful design.
Control Volume
Qflow Qinput Qwalls Qspe
Qpump Qfan
Qmask
Fig. 4: Calorimeter energy balance for standard test procedures.
6.0 NOMENCLATURE variables: A Area (m2) Specific Heat (J⋅kg-1K-1) Cρ ε Emissivity (dimensionless) F Solar Heat Gain Coefficient (dimensionless) G Irradiation level (W⋅m-2) P Power (W) Q Energy Flux (W) R R-value (m⋅K⋅W-1) ΔT Temperature Difference (oC) η Efficiency (dimensionless) τ Time Constant (sec)
subscripts: f fenestration fan calorimeter circulating fan flow flow loop input solar input mask mask wall pump calorimeter circulating pump ref reference heater walls calorimeter walls
7.0 REFERENCES [1] van Wonderen, S.J., and Harrison, S.J. (1992), “Development of an Improved Calorimeter Cell for Window Thermal and Solar Performance.” Presented at Solar Energy Society of Canada Inc., Edmonton, Canada. [2] Collins, M.R., and Harrison, S.J. (1998), “Determination of the Inward-Flowing Fraction of Absorbed Solar Energy for Venetian Blinds.” Presented at RETSSS’98, Solar Energy Society of Canada Inc., Montreal, Canada. [3] Harrison, S.J., and Bernier, M.A. (1984), “A Reference Heat Source for Solar Collector Thermal Testing.” ASME publication 84-WA/Sol-2. [4] SCL (1993), “QMON Manual: Solar Calorimetry Laboratory.” Queen’s University at Kingston, Ontario, Canada. [5] Harrison, S.J., and Barakat, S.A. (1983) “A Method of Comparing the Thermal Performance of Windows.” ASHRAE Transactions, Vol. 89 (1).
