The installation of pyranometers, pyrheliom eters and pyrgeom eters is relatively sim ple (Annex B provides inform ation on som e of the instrum ents that m ay be suitable for use at BSRN stations), but nevertheless requires care and attention to detail.
Originally, the BSRN recom m ended that the m anufacturer and type of instrum ent used for the m easurem ent of global radiation should also be used for the m easurem ent of diffuse radiation. This was to reduce uncertainties associated with the tem perature and angular responsivities of the instrum entation. In consideration of the uncertainty associated with the therm al offset associated with
‘black’ therm opile instrum ents, high quality B&W instrum ents, that have low therm al offsets, have identical, or a sim ilar dom e spectral characteristics (particularly in the shorter wavelength range), and have good directional responsivity m ay be substituted. A careful analysis should be perform ed to determ ine that the m eans of m easuring both diffuse and global irradiance m inim izes the overall uncertainty of the m easurem ent.
It was also recom m ended originally that the pyrheliom eter used be of the sam e m anufacturer as the pyranom eters, prim arily to ensure that the spectral response of the two instrum ents be identical.
Subsequently, it was discovered that the use of instrum entation constructed by the sam e m anufacturer did not guarantee sim ilar spectral responses because different m aterials were used for the pyranom eter dom es and for the optical flats of the pyrheliom eters. Rayleigh scattering alters the spectral distribution of energy at the surface so that the proportion of near infrared (NIR) radiation in the diffuse irradiance signal is less than that m easured in the direct beam and in the global radiation spectra. To capture this radiation signal, which can be significant in a dry atm osphere, the suggestion has been brought forward to use quartz, sapphire or calcium fluoride optical flats on pyrheliom eters. The m anufacture of dom es using the latter two m aterials is uncom m on, expensive and unnecessary for the m easurem ent of diffuse irradiance. Therefore, the selection of the types of instrum ents to be used for the m easurem ent of the various com ponents m ust be based on a careful uncertainty analysis of the instrum entation and the atm osphere under which the m easurem ents are to be m ade. Included in this analysis m ust be the recognition that the three m easurem ents are to be used in the quality assurance of the data.
A num ber of docum ents, including m anuals provided by m anufacturers, have been published that include inform ation on the installation of these instrum ents. Docum ents published by technical agencies include:
Radiation Measurement. International Field Year for the Great Lakes, Technical Manual Series No.
2, National Research Council of Canada, 1972.
Revised Instruction Manual on Radiation Instruments and Measurements. W orld Clim ate Research Program m e, W CRP Publication Series No. 7, W MO/TD No. 149, 1986.
Meteorological measurements concerning questions of air pollution, Global radiation, direct solar radiation and net total radiation. VDI-Richtlinien, VDI 3786, Part 5, 1986.
Solar Energy - Field Pyranometers - Recommended practice for use. International Standards Organization Technical Report TR9901, 1990.
4.2 Installation of pyranometers and pyrgeometers 4.2.1 Pre-installation Checks and Service
Before installing any pyranom eter the instrum ent should be carefully inspected.
(1) If not provided by the m anufacturer, the instrum ent should be calibrated so that the following inform ation is available:
(i) the responsivity of the instrum ent to radiation (ii) the spectral range of the instrum ent
(iii) the linearity of the instrum ent between 0 and 1500 W m-2
(iv) the directional responsivity of the instrum ent (cosine and azim uthal response of the instrum ent) for pyranom eters
(v) the deviation of the tem perature com pensation circuit of the instrum ent over the tem perature range (-10° to +40° of range) or if not com pensated, the required tem perature correction of the instrum ent. In clim ates where the tem perature range is greater than that specified, instrum entation should be selected to m eet the tem perature regim e.
For the m ost accurate m easurem ents, the tem perature of the therm opile should be m onitored and the signal corrected for tem perature variations. A num ber of instrum ents using non-therm opile sensors m ay be considered if they m eet all other criteria.
(vi) the instrum ent has been radiom etrically levelled. That is, the therm opile is horizontal when the bubble level indicates such (the bubble level should have an accuracy of
±0.1°).
(2) Checks should be m ade of all wiring to ensure that there are no nicks in the sheathing nor stress on the connections. The wire should be of a variety that will withstand the clim atic regim e of the area in which the instrum ent is to be installed.
(3) All O-rings should be lubricated lightly with a very fine grease (e.g. Dow Corning Model 55 O-ring lubricant or Fischer Scientific Cello-Seal C-601).
(4) All threaded parts should be lubricated in a m anner sim ilar to the O-rings.
(5) The therm opile (or the signal transducer) should be visually inspected to ensure that the surface is uniform in colour and texture.
(6) The BSRN accuracy guide indicates that the case tem perature of the instrum ent should be m onitored. If the instrum ent is fitted with a therm al m easuring device, the wiring should be checked and the reduction algorithm tested at known tem peratures. In the case of pyrgeom eters, all therm istors should be tested.
(7) The inner and outer dom es should be checked for scratches or nicks. If found, the dom es should be replaced. In the case of pyrgeom eters a sim ilar check should be m ade of the silicon dom e. However, if the dom e requires replacem ent due to dam age, the instrum ent m ust be re-calibrated.
(8) The im pedance of the instrum ent should be checked against the m anufacturer’s values.
(9) The desiccant should be fully activated. It is recom m ended that the desiccating m aterial be of the bead type (e.g. Trockenperlen, Kali-chem ie AG) and not one which easily powders (e.g. Drierite, Ham m ond Drierite)
(10) All connectors m ust be waterproof and should be appropriate for the clim atic conditions in which the sensor will be deployed. For exam ple, in m arine environm ents care m ust be taken against using connectors that are prone to corrode. It is recom m ended that keyed connectors be used for greater safety in m aintaining instrum ent polarity.
4.2.2 Mechanical Installation
(1) The instrum ent should be m ounted with the direction of the connector facing poleward for fixed platform s and away from the solar disk when m ounted on solar tracking devices.
(2) The instrum ent m ust be fastened to the platform (or ventilating device (see below)) so that it will not m ove in inclem ent weather. The bolts used should be lubricated before assem bly for ease of disassem bly. Initially, these bolts (norm ally two or three depending upon the instrum ent) should be not be tightened until the instrum ent is levelled according to its bubble level.
Figure 4.1. Ventilator with m otor located beside the instrum ent as used by Deutscher W etterdienst.
Spring loaded bolting devices for m ounting the instrum ent are also an excellent m eans of guaranteeing the instrum ent will rem ain fixed while providing the added ability of levelling the instrum ent without requiring the bolts being loosened.
(3) The instrum ent should be levelled using the supplied three levelling feet. By first adjusting the foot closest to the bubble, the instrum ent should be adjusted until the bubble is centred within the inner circle of the supplied bubble. W hen com pletely centred, and radiom etrically levelled, the bubble level indicates that the therm opile is horizontal to within ±0.1° causing an azim uthal variation of ±1% at a solar elevation of 10°.
(4) Carefully tighten the retaining screws so that the instrum ent is im m ovable. To do so, gently tighten the bolts alternately until secure. Be careful not to over-tighten.
(5) Place and adjust the radiation shield or ventilated housing cover so that it is parallel to, and level with or below the therm opile surface.
4.2.2.1 Ventilated housing
The recom m ended procedures for the m easurem ent of global radiation require the use of a ventilated housing to im prove the overall stability of pyranom eter m easurem ent by dam ping changes in the pyranom eter body tem perature due to solar loading and potentially reducing the therm al offset. In som e clim ates, the use of a ventilator also im proves the am ount of recoverable data by elim inating dew and reducing the num ber of occurrences of frost and snow on the instrum ent dom es. Measurem ents in other regions, however, have not shown a significant increase in accuracy or percent data recovered with the use of ventilated housings. As each ventilator adds extra cost and com plexity to the installation and m aintenance of a station a thorough analysis of its requirem ent should be m ade before installation.
Locations where a ventilated housing are recom m ended are:
(1) where dew, frost or snow is prevalent,
(2) where natural ventilation is infrequent or variable,
(3) where there is significant radiative cooling during portions of the year, a ventilated housing m ay reduce therm al-offset,
(4) where the hum idity is high during portions of the year a ventilator will reduce the possibility of water dam age and reduce the frequency of desiccant changes.
Figure 4.2. Ventilator with the m otor located beneath the instrum ent. Note the extra ventilation holes near the top of the housing used to reduce snow accum ulation (Davos, Switzerland).
The two recom m ended styles of ventilated housing are:
(1) W here the ventilator fan is situated beside the instrum ent and the pyranom eter is com pletely enclosed so that the air flows evenly around the dom e. Figure 4.1 illustrates this type of blower as used by the Deutscher W etterdienst. The advantage of this design is the ease in which a fan can be replaced without tam pering with the pyranom eter. Conversely, the design is m ore com plex because the air is entering from one side of the pyranom eter and m ust be funnelled around the instrum ent to pass over the dom e equally from all directions. This m ay require the use of a larger fan than those ventilators that pass air around the instrum ent from beneath. The tem perature of the instrum ent in the encapsulated ventilator and, to a lesser extent the instrum ent dom e, will rise slightly above the am bient tem perature due to the heating of the air by the blower m otor.
(2) W here the housing encloses the pyranom eter and the ventilator fan is located beneath the instrum ent and blows air from beneath the housing, around the instrum ent and over the dom e (Figure 4.2). This is the m ore com m on of the two recom m ended ventilation system s. The power dissipation heats the incom ing air by approxim ately 1°C, which in turn heats the body and dom e of the enclosed instrum ent. Unlike (1), the instrum ent m ust be rem oved from the housing before a fan can be replaced. The area beneath the instrum ent m ust also be kept free from obstructions to m aintain airflow.
The heating effect on the dom e due to the fan m otors is negated when wind speeds are m oderate to high.
Heating resistors can be added to both ventilators if required during cold weather ope rations. Care m ust be taken, however, in that these m ay also alter the overall response of the instrum ent.
Figure 4.3. An one-axis tracker used in shading a pyranom eter. Note the use of two fine wires to m aintain the stability of the shading disk. (Developed by
Deutscher W etterdienst).
4.2.3 Mechanical installation of shaded sensors (pyranom eters and pyrgeom eters)
The general installation of shaded sensors follows the guidelines set out in 4.2.2, but includes the added com plexity of aligning the shading device with the instrum ent. Within the BSRN, shade rings (diffusographs) are not accepted as a m eans of shading an instrum ent because of the field of view the ring subtends.
Two types of devices are com m only em ployed within the BSRN.
Figures 4.3 and 4.4 illustrate synchronous m otor shade devices as designed by the Germ ans and the Swiss. The Germ an device uses a single axis to carry the pyranom eter about which the shade-disk rotates. This design is significantly m ore com plex than the Swiss design, but is m ore efficient in using space. The Swiss design fixes the pyranom eter on a stable, level stand m ounted separately from the equatorial m otor. The position of the shade disk m ust be m oved along the shading arm with the changing solar declination. The size of the arm holding the disk m ay reduce the am ount of diffuse irradiance detected by the sensor. To reduce this effect, the Germ an design has reduced the overall dim ensions of the shade-arm by adding guy-wires to the design. Consideration m ust be given to the location of the instrum entation when determ ining how robust the design and dim ensions of the shade-arm need be. Locations that experience high winds, driving precipitation or significant snowfalls will require a m ore robust design than less hostile environm ents.
W hile inexpensive to fabricate, these devices require m ore daily m aintenance than shade system s that operate on two-axes trackers. Problem s associated with synchronous m otor shade arm s are sim ilar to those of solar trackers used for the m easurem ent of direct beam irradiance (Section 4.4).
The general installation of devices sim ilar to those of the Germ an and Swiss that use a single-axis synchronous m otor system for shading are equatorial m ount are sim ilar.
(1) The pyranom eter be m aintained in a stable, horizontal position.
(2) The synchronous m otor m ust:
(I) be wired appropriately the electrical power frequency of the location of installation, (ii) be wired to follow the path of the sun (opposite for northern and southern hem ispheres, (iii) be m ounted on a level baseplate,
(iv) be perpendicular to the horizon,
(v) be located equatorward of the pyranom eter, (vi) be aligned true north-south,
(vii) be pointing poleward at an angle equal to the latitude of the site,
(viii) be in a position such that a line extending from the m otor axis pass through the centre point of the instrum ent sensor.
(3) The cables of the pyranom eter (and ventilator) m ust be routed in a m anner that will not interfere with the shade arm . In the Swiss case, this is done by having the shade arm able to rotate about the stand on which the pyranom eter is located, while in the Germ an case the wires are passed through the centre of the axis.
(4) The area around the instrum ent m ust be free of obstructions so that the arm can rotate a full 360°. This includes the surface on which the instrum entation is m ounted.
Two-axes trackers can be used either as a m eans of shading one or m ore instrum ents or as a com bination unit where a pyrheliom eter is also attached to the tracker. Figure 4.5 illustrates the Australian com puter-controlled active-tracker on which the norm al incident direct beam is m easured using a pyrheliom eter coincidentally with a pyranom eter being shaded with a shade-disk, while Figure 4.6 shows a Canadian
Figure 4.4. View of two Swiss oversize tracking disks. Note how the pyranom eter is physically separated from the m otor and the shade device. The increased width of the arm holding the shade disk elim inates the need for stablilizing wires, but increases the am ount of sky obscurred. The slot along the arm is for the m ovem ent of the shade disk. On the instrum ent to the right of the photograph, the vertical wires are used to deter birds perching on the instrum ent. (Courtesy of MeteoSwiss, Payerne, Switzerland.)
M ajor, G ., 1992: Estim ation of the error caused by the circ um solar radiation when m easuring global
12
radiation as a sum of direct and diffuse radiation. Solar Energy, 48(4), 249-252.
designed system shading both a pyranom eter and pyrgeom eter along with m easuring the norm al incident direct beam (see Section 4.4 for details on two-axes trackers). To use the tracker as a platform for both the shading of a pyranom eter and the pointing of a pyrheliom eter, the elevation drive m ust be m echanically translated so that it is horizontal and at the sam e height as the signal transducer of the instrum ent to be shaded. In both cases, the shade disk(s) (or the shade sphere(s)) rem ain at a fixed distance from the instrum ent sensor via the cantilevering-m otion provided by the arm ature. Once installed, care m ust be taken not to rotate the elevation drive of the tracker below the position where the cantilever system binds on itself. This is accom plished by not allowing the elevation axis to go m ore than about 5 - 10° below the horizon before program m ing the tracker to “go to sleep” and/or return to a pre-sunrise position. Depending upon the type of tracker (especially if it is a friction drive), the num ber of steps taken to m ove 360° m ust be checked. This m aintenance procedure requires that the shade assem bly be disconnected from the tracker.
W hether using a single-axis or two-axes tracker, the instrum ent requiring shading m ust be placed preciselyso that the shade of the diffusing disk com pletely covers the outer dom e of the instrum ent.
The general rule presented by the W MO for pyrheliom etry is that the ratio between the length and the diam eter of the opening angle of a pyrheliom eter is 10:1. This rule can be used to approxim ate the geom etry of the disk or sphere used to block the irradiance from solar disk. Major (1992) discusses12 the use of pyrheliom eters and shaded pyranom eters for calculating global radiation with respect to the optim um design of the shade disk. The results indicate that the best equivalence can be expected if the distance between the receiver and the shading disk is chosen so that the slope angle is larger and the opening angle is less than those of the pyrheliom eter in use. Major (Personal Communication) has calculated that diffuse/direct irradiance m easurem ents m ade at various BSRN stations using standard system s m ay increase discrepancies between the global irradiance m easured by a pyranom eter and the sum m ation of the diffuse and direct irradiances by up to 5 W m . Optim ized arm lengths and shade-2 diam eters im prove this to better than 0.5 W m . Table 4.1 provides optim al geom etry for several com m on-2 com binations of pyranom eters and pyrheliom eter or cavity radiom eter. Further work on this issue has been done by the BSR N W orking Group on Diffuse Geom etry, culm inating in a final report that is reproduced as Annex C.W hile the advantages of using a two-axes tracking system for m easuring both diffuse and direct beam irradiance are num erous, there are several disadvantages that should be
The general rule presented by the W MO for pyrheliom etry is that the ratio between the length and the diam eter of the opening angle of a pyrheliom eter is 10:1. This rule can be used to approxim ate the geom etry of the disk or sphere used to block the irradiance from solar disk. Major (1992) discusses12 the use of pyrheliom eters and shaded pyranom eters for calculating global radiation with respect to the optim um design of the shade disk. The results indicate that the best equivalence can be expected if the distance between the receiver and the shading disk is chosen so that the slope angle is larger and the opening angle is less than those of the pyrheliom eter in use. Major (Personal Communication) has calculated that diffuse/direct irradiance m easurem ents m ade at various BSRN stations using standard system s m ay increase discrepancies between the global irradiance m easured by a pyranom eter and the sum m ation of the diffuse and direct irradiances by up to 5 W m . Optim ized arm lengths and shade-2 diam eters im prove this to better than 0.5 W m . Table 4.1 provides optim al geom etry for several com m on-2 com binations of pyranom eters and pyrheliom eter or cavity radiom eter. Further work on this issue has been done by the BSR N W orking Group on Diffuse Geom etry, culm inating in a final report that is reproduced as Annex C.W hile the advantages of using a two-axes tracking system for m easuring both diffuse and direct beam irradiance are num erous, there are several disadvantages that should be