This section describes the routine operation of the 8-port smoke and SO₂ sampler. Procedures for the weekly changeover of bottles and filter papers are outlined. This is followed by further details on specific aspects of the sampler’s operation and routine maintenance, such as obtaining clear smoke stains, reading the gas meter and maintenance of the pump.

The 8-port sampler is designed with eight pairs of clamps and bottles for weekly operation, providing daily means on a midnight to midnight basis. The timed eight-port valve should be set to switch over at midnight, to expose a fresh paper and bubbler each day. Weekly visits are made to change solutions and filter papers. The exposed solutions and papers are taken back to the laboratory for analysis and measurement.

The visits should ideally be made on the same day each week. If the sampler is left unattended for more than eight full days, filters and solutions will be exposed twice, invalidating two days’ data for each day this happens. It is not necessary to ensure that the visit takes place at a particular time of day; the data processing programs used by AEA Technology Environment accommodate meter readings at varying times. The eight-port valve (if set correctly) ensures that the bubblers change over at midnight GMT exactly.

A standard data return form is used. These pre-printed forms are issued monthly to site operators by AEA Technology Environment. Section 8 gives full details of this form. Operators should take the data return form to the site when they visit, and record the meter reading and other details directly onto the form, adding titration and reflectometer results later. Recording directly onto the form in this way avoids transcription errors. If this is not possible, a log-book should be used to record meter readings, times, titration results, reflectometer readings and any notes. The log-book should be kept for this purpose only, and all entries should be clearly identified with the date, and site. Do not use loose sheets of paper or rough notes; these are liable to get lost.

Each pair of clamps and the corresponding Drechsel bottle head should be labelled with the number 1 to 8, in the order in which they operate. This should correspond to the number of the port on the 8-port valve to which it is attached (see back of 8-port valve). The Drechsel bottle heads and filter clamps can then be numbered to correspond. Suitable marking materials include wax "chinagraph" pencils, waterproof felt-tip pens or markers, or self-adhesive labels providing they stick firmly to glass or metal.

A suggested weekly procedure is given below. It is expected that operators will adapt this to some extent; however, it is important the procedure remains consistent from week to week, to avoid errors.

It is useful to have a marker, eg. a coloured clothes peg, which is placed each week on the head of the bottle which is active on the day of the visit (not on the tubing). This bottle and filter will be left undisturbed on this occasion, to be changed at the next visit; the marker should then be transferred to the active bottle. When changing solutions and papers, it is recommended to start with the bubbler and clamp which were active yesterday (for example, if bottle 5 is bubbling today, start with bottle 4), and work backwards round the circuit. The procedure will vary slightly depending on whether the operator uses a second set of Drechsel bottles to transport the exposed samples, or uses bottles of another type.

A card containing these instructions should be left with the apparatus so that, in an emergency, someone unfamiliar with the method could carry out the procedure.

1. There is no need to switch off electricity supply to pump in the case of the 8-port sampler, as the active bubbler will not be disturbed.

2. Ensure the timer is showing the correct time with respect to the next changeover. If a power cut has occurred, and the interruption was less than two hours, the timer should be re-set but the data need not be discarded. If the interruption was longer, inform AEA Technology Environment and make a note on data return form.

3. Read the meter. Enter figures direct on record form against today’s date, ensuring that the pre-printed decimal point is in the correct place. Also enter size of filter clamp in use, and time of meter reading to the nearest minute. Do not make any corrections for British Summer Time.

4. Now check the flowrate: subtract previous meter reading from present one and calculate the volume of air passed since the last visit. If it is outside the required range of 2.0 cubic metres per day, + 10%, this will need to be investigated and any possible fault corrected. It is important that the flowrate is checked each time the site is visited. If the flowrate is outside the acceptable range, investigate the cause immediately.

5. When changing the solutions and filters, work backwards in numerical order from today’s active bubbler (which is left alone). The last bottle to be changed should carry the marker from last week’s visit, and the date should fit into the sequence.

6. Ease the first bubbler out of its holder . Take the bubbler head firmly with one hand (without touching the glass tubes) and ease off the bottle with the other .

7. Either ( i) replace with a second bubbler bottle containing fresh 1 vol hydrogen peroxide solution, seal the bottle containing the exposed solution, ensuring it is marked with the site name and date of exposure or (ii) pour contents into spare bottle or container for transport, and re-fill bottle with fresh solution to level marked on bubbler. See that container or spare bottle is named, dated and firmly sealed.

8. Do this for all seven non-active bottles. As you replace the bottle, ensure that the bubbler head makes a good seal. (If spare containers are used for transport, they must be conditioned before use as described in section 6). When not in use they must be kept filled with fresh, pH 4.5 hydrogen peroxide solution which is emptied out on the morning of the day they are used, several seconds being allowed for drainage).

9. Unscrew filter clamp and withdraw paper. Slip fresh sheet into place, making sure that smoother side is placed downwards, facing incoming air. After checking that the two halves of the clamp are absolutely in line, tighten screw down hard.

10. Carefully write the date of exposure, and the site name or number on the edge of the paper which was removed. Use a pencil or ball point pen which will not smudge. Do not touch the stain itself, as this could affect the result. Place paper in a self-sealing sample bag, box or envelope. Do not let the stain side of two exposed papers touch, as dirt may transfer from one to another. Note: some operators prefer to mark the date and site on the papers before inserting them in the clamps.

11. Do this for all seven non-active bottles and clamps.

12. Ensure that the sampler is bubbling normally before you leave. Note the rate of normal bubble formation; if subsequently the air flow is suspect, the bubble rate may help determine whether the pump or the meter is at fault.

The operator should now have seven exposed peroxide solutions, and seven smoke-stained filter papers.

If the timing disc is not in the expected position when a visit is made and/or bubbling is taking place in the wrong bottle, then a power cut has probably occurred. Provided that the interruption does not exceed 2 hours, no further action is necessary other than to re-set the timing disc. If the interruption exceeds 2 hours, then some of the results may have to be rejected. A note should be made on the data return form and/or in the log book of the time and duration of the interruption to the electricity supply, for quality control purposes and the timing disc will require re-setting. Should this persistently occur, the problem may not be power cuts, but a faulty timer.

The exposed solutions must be carried back to the laboratory for titration in securely sealed containers.

The best option is to keep two sets of Drechsel bottles for the sampler; each week, the bottles containing the exposed samples can be marked with the exposure date, removed from the sampler and replaced with the other set filled with fresh hydrogen peroxide solution. They are strong, easily "conditioned" and can be fitted with a stopper of an inert material such as glass. The exposed solutions remain in the same Drechsel bottle until titrated, minimising the possibility of contamination. However, Drechsel bottles are comparatively expensive.

A cheaper option is to use a set of screw capped inert plastic or glass bottles, for transport and storage of exposed solutions until they can be titrated. Suitable bottles can be obtained from a laboratory supplier. They should be tested as described in section 5.3.2. , to ensure they do not affect the pH of the solutions.

5.2 OBTAINING GOOD SMOKE STAINS

Always use Whatman Grade 1 filter paper. This is available from the major laboratory suppliers, and is not expensive. Several sizes are available:

• Whatman reference 1001 042 is a 42.5 mm diameter circle. The brass filter clamps in an AGL model 8-port sampler are also 42.5 mm diameter; however this can be awkward in use because once the paper is in the clamps, there are no protruding edges for handling or marking the paper.

• Whatman reference 1001 055 is a 55 mm diameter circle. Ideal for the Airtest (Glass Developments) and Eccodata samplers, there is plenty of gripping and writing space. However, this size does not fit the AGL model well, as the paper is too large to fit between the clamp alignment/centring posts, leaving a crescent shaped space in the clamping area, which may leak.

• Whatman reference 1001 824 is a rectangular paper 75 mm x 100 mm. This size would be wastefully large if used in one piece, but by guillotining through the long edge to give two 75 mm x 50 mm pieces it can be used in all 1 inch clamps. This is ideal in the AGL model when inserted diagonally.

The filter papers have a rough side and a smooth side. They are all the same way up in the pack or box. Determine which side is uppermost and write on the outside of the box for future reference. Take care when guillotining rectangular papers not to turn the pile over: keep them all the same way up.

When inserting into the clamps, the smooth side should face the incoming airflow, and collect the smoke stain. When reading the smoke stains, use the same side of a clean paper from the same batch to set the '100' reading. Use a fresh paper for this every measurement session.

Smoke stains should be perfectly circular, and have sharp edges with no blurring or rings around them. Fig. 5.1 shows examples of good and bad smoke stains. To obtain good smoke stains:

1. The lower part of the clamp must be correctly positioned. The means of positioning the clamp depends on the sampler design. In the AGL sampler this is done by pushing it back until it meets two ‘centring posts’; some older designs have a small metal protrusion on the floor of the case, which fits into an indentation on the underside of the lower part of the clamp.
2. The upper and lower filter clamps must be exactly aligned, or the resulting stain will not be circular, the edges will be ill-defined, and the concentration of smoke obtained from it will be inaccurate, possibly by as much as 40%.
3. The clamps must be fully tightened, otherwise indoor air will leak in between them. This will produce stains with fuzzy edges, and dilute the sample, giving inaccurate results.
4. The paper should cover the whole area between the clamps; otherwise there may be leakage.
5. The filter paper must always be inserted with the smooth side facing the incoming air, normally downwards. To tell which is the smooth side of the sheet, the paper should be held with one edge towards the light, when the rough side can easily be seen as the light passes across the surface.
6. The filter papers which have been removed from the clamps should have the date and the name of the site written on them without touching the stain, and it should be placed in a clean sample bag or box until required. Alternatively, it may be helpful to mark the site identification and exposure date on the papers before they are inserted into the clamps.
7. One of the most common errors is reading the stains on the wrong side: it is therefore important that a consistent system is adopted, such as marking the paper on the same side as the stain.
8. If dark stains (reflectometer reading less than 70) are occurring frequently, it may be necessary to switch to a larger clamp size. However, 1 inch clamps are sufficient for most applications in the UK nowadays, so if this is a problem please consult AEA Technology Environment for advice.

The faces of the clamps may become worn after a long period of use. To improve the surfaces of brass clamps, place a sheet of carborundum paper on a perfectly flat surface such as a sheet of plate glass and rub the face of each clamp over it with a circular motion; carborundum paper grades 0, 00 and 000 should be used in succession as necessary. Test the clamps again and repeat until the seal is restored. The clamps must be carefully cleaned after this treatment. Alternatively, return them to the supplier for re-surfacing.

5.3 BUBBLER (DRESCHEL) BOTTLES

5.3.1 Use and Care of Bubblers

The Drechsel bottle, or bubbler, is made in two parts: the body or bottle itself, and the head. To empty and re-fill, the bottle is disengaged from its head, which remains permanently in position. To do this, release the Drechsel bottle from any clips etc. holding it in place in the sampler. Then hold the Drechsel head firmly by the head itself (not by the glass tubing) with one hand while twisting the bottle downwards with the other. Any liquid remaining in the inlet tube is removed by touching it against the inside of the bottle, not the inside of the ground glass joint.

The ground glass joint between the Drechsel head and the bottle, under clean conditions, with careful use, will be airtight. A very small amount of tap lubricant, as used for burette taps, may be needed if the joint sticks.

New Drechsel bottles may contain traces of alkali, which would interfere with the measurement of sulphur dioxide. They may therefore need to be "conditioned" before use, in order to remove this.

The new Drechsel bottle (or other glass container for transporting exposed samples if used) should be washed with deionised water, and rinsed well with 1 vol pH 4.5 hydrogen peroxide solution. Fill with the peroxide and seal with the stopper (or screw cap) provided. Leave until next day. Then pour around 50 ml of the peroxide from the bottle into a flask or beaker, and add 3 to 5 drops of BDH pH 4.5 indicator. Compare the colour with that of the same volume of fresh solution, in a similar beaker, to which indicator has also been added. If the colours of the two samples are identical, there has been no change in the solution left in the container overnight, and the container is suitable for use. If the sample from the new bottle is at all blue, the bottle should be refilled with fresh dilute pH 4.5 peroxide, and left overnight again. This procedure should be repeated until there is no colour change.

Some water will evaporate from the bubblers while in use. This in itself is not a problem, as long as there is sufficient liquid to keep the end of the inlet tube submerged by 1 cm throughout the sampling period. The volume is not critical to the subsequent titration. In the UK, an initial volume of about 50 ml of peroxide per bottle is normally sufficient. However, in warm locations, the volume of hydrogen peroxide solution should be increased (though not to the extent that any liquid could be drawn into the outlet tubing). It may sometimes be possible to minimise the heating by leaving the box open, as the pump generates heat and the inside of the box can become quite warm.

Freezing is much less likely in an 8-port than in the old single-day samplers, as the pump, which is a good source of heat, is enclosed within the box. However, if freezing does occur it will be necessary to insulate the entire sampler. The meter must be kept at least as warm as the bubbler or condensation will take place within it.

Under no circumstances should any form of "anti-freeze" be added to the solution as this would result in an incorrect answer being obtained on titration.

5.4 THE 8-PORT VALVE

The 8-port valve, by means of its timer, can be set to operate at any time of the day. However, for the purposes of the Network it should be set to change over at midnight GMT, throughout the year. Once the timer is set, it should run without adjustment for many weeks, assuming there are no power cuts or other problems. Instructions for setting the most commonly used 8-port valves (those found in the AGL, GDL, and Eccodata samplers) are given below. Instructions for the 8-port valves with the digital (Randall) timer are given in Appendix 2.

Usually, the only part of the valve to need adjustment is the timing disc. To move the disc, the locknut in the centre must be loosened first; forcing the timing disc round with the locknut tightened could cause damage. Hold the timing disc still with the fingers of one hand and loosen the locknut with the other hand. Rotate the timing disc to the new position, hold it in place and screw the locknut tight again - hand tight is sufficient. To test that the locknut is tight, try to gently move the timing disc by pressing with one finger. (The Eccodata model does not have this locking mechanism).

Avoid always moving the timing disc in one direction; constant rotation in one direction will either lock the disc, or wind it so far out that the microswitch inside is no longer operated by the cam on the timing disc. It should never be necessary to turn it through more than one revolution in either direction.

The direction of rotation of the timing disc has not been standardised. The hour markings on the timing disc may indicate either the time which has elapsed since the last change (in this case, the disc rotates past the timing mark in the direction 0-22-20-18 ...), or the time until the next change (in which case the disc rotates in the direction 0-2-4-6 ...). As in the twenty-four hour clock, 0 hours and 24 hours are at the same point on the disc, which for convenience is marked 0. If the valve is required to change from one port to the next at a specific time each day (usually midnight), the timing disc must be set correctly.

( i) If the timer shows hours until next change; check the time now (eg. 2.30 pm). If it is British Summer Time (BST), subtract 1 hour (eg. to 1.30 pm GMT). Calculate the hours remaining until midnight GMT (10 hours 30 minutes). Loosen the locknut, set the dial to show a time of 10.30 against the datum mark, and relock the disc.

(ii) If the timer shows hours since last change; the dial effectively shows the time of day. Check the time now (eg. 10.00 am). Again, if it is British Summer Time (BST), subtract 1 hour (eg. to 09.00 am GMT). Loosen the locknut, set the dial to show 9.00 against the datum mark, and relock the disc. There is no need to alter the timer when British Summer Time begins or ends.

If it is necessary to select a different sampling line from the one in use, the port-change motor can be operated manually as follows.

1. Have the sampler switched on and bubbling.
2. Hold the timing disc still by gentle pressure from the fingers, and loosen the knurled nut (locknut) by a few turns with the other hand.
3. Rotate the timing disc gently, in either direction, until the port-change light comes on, indicating that microswitch inside the valve is connected. This happens when the 0 on the disk is lined up with the mark on the body of the valve (in practice the two marks may not quite coincide at this point). The opening of the microswitch can be felt in the fingers rotating the timing disc - there is a distinct change in resistance to the turning of the disc. Now wait.
4. The bottle in use will stop bubbling. After a second or two, the bubbler next in line will start.
5. Wait until the light goes off.
6. If it is necessary to move on more than one port, repeat actions 3 to 5 above.
7. Set the changeover time to midnight GMT as described above, and re-tighten the locknut holding the timing disk in place.

It is also possible to set the changeover time by setting the dial so the datum mark is on zero, allowing the port change light to come on and bubbling to stop. Then switch off the sampler completely until midnight GMT. This method is more accurate, as it avoids error due to any ‘backlash’ in the valve’s gears ; however, it is of course very inconvenient, so in practice most site operators use the method above, which is perfectly adequate for the purposes of the Network.

Although the 8-port valve requires very little maintenance, it may leak if its internal components become worn. It is therefore recommended that it is serviced from time to time (and certainly if a leak is suspected). The current suppliers of samplers are able to provide this service, for any type of 8-port.

5.5 READING THE GAS METER

Mis-reading the meter, or mis-recording the reading, is a common source of error. Operators are advised to take their time and check their readings carefully.

Always note the time of the meter reading, correct to the nearest minute. The time should be recorded in 24 hour clock form, eg. 9 am as 0900, 1.30 pm as 1330, and so on. Record the actual date on which the meter reading was taken: do not worry if the meter reading was not taken on the "normal" day of the week.

A variety of different types of gas meter are now in use. The more recent AGL and Airtest models use the metric Remus gas meter; this is a Spanish design, and uses a comma to represent the decimal point. Other types are in use, particularly in older samplers. Appendix 1 gives further guidance on reading various types of meters. The entry on the Data Return Form should be entered to the nearest 0.01 m³ for a metric meter, or to the mearest 0.1 ft³ for one reading in cubic feet: it is not necessary to round up the reading. Leading zeros must appear where appropriate, eg. 02.34. You may find it helpful to cover part of the meter display window with masking tape , so that only the four required digits are visible. It may also be helpful to mark the position of the decimal point.

The accuracy of the meter should be within +/- 3%. AEA Technology Environment can test your meter free of charge to determine whether or not it meets this specification. Meter suppliers may also offer this service. If there is any doubt about the accuracy of a meter, or if it has been in use for several years without checking, it is recommended that the meter should be tested.

Each time the meter is read, the volume of air passed since the last reading should be calculated. Please check air volume flow-rates are within the acceptable range of 1.8 to 2.2 m³ per day (65 to 75 cubic feet per day). This is approximately 14 m³ (500 cubic feet) per week. If this check is done while still at the site, mis-readings can be corrected, and faults identified quickly. Any unexpected variation should be checked; this way, any problems with the apparatus can often be remedied before they become serious and lead to loss of data. If a power cut is suspected, please indicate this on the Data Return Form: on which day and for how long, if known.

If someone other than the usual person has to do the meter reading, please make sure they are instructed, and given a copy of this manual. Instructions can also be kept near the sampler.

5.6 CARE AND MAINTENANCE OF THE PUMP

5.6.1 The CAPEX Pump

The type of pump now recommended for use in the 8-port sampler is the Capex 2LNS, manufactured by Charles Austen Pumps Ltd, a small diaphragm pump. This pump is capable of producing flowrates of over 7 m³ per day; the flowrate is controlled to within the required range (2 m³ per day + 10%) by a flow limiting critical orifice.

The critical orifice assembly is fitted between the meter outlet and the pump inlet; it consists of a sealed black plastic cylinder about 5cm long, housing a 5µm filter (the "safety filter"), which captures any remaining dirt particles in the sampled air. This cylinder has a small side boss; threaded into it is a small white connector which contains a flow limiting orifice itself - a tiny disc of hard material with a hole in its centre. Any damage or blockage of this component will adversely affect the flowrate; the purpose of the safety filter is to capture any particles remaining in the sampled air, which would otherwise build up on the critical orifice, and eventually block it. This assembly is connected to the inlet of the pump itself by a 20 cm length of reinforced tubing (ordinary PVC tubing would collapse under the pump suction). On the pump outlet there is another black plastic cylinder similar to the safety filter case; this is a silencer (and is marked as such).

The manufacturer, Charles Austen Pumps Ltd, recommends that the whole filter and critical orifice assembly should be replaced every two years. If the flowrate is checked each week, any persistent downward trends in flowrate should be spotted before the flowrate falls below the lower limit of 1.80 m³ per day, giving ample time for a new filter and orifice assembly to be fitted. We recommend that you keep a new filter and critical orifice assembly to hand . This way, replacement can be carried out quickly and easily, without losing any data. They are available from suppliers of Capex pumps (see our list of suppliers), are not expensive, and are supplied ready assembled, with the filter case sealed and the correct length of reinforced tubing. Do not fit a new critical orifice without also replacing the filter, or the new orifice will block very quickly.

Replacement pump diaphragms are also available from the manufacturer, or from the supplier of the sampler.

Although the CAPEX is now the recommended pump, some samplers still contain the older Dymax Mk IIA pump. This was designed to sample at the standard rate without attention for a year. At least once a year it should be returned to the makers for servicing, even if it appears to be running correctly. An increase in the amount of noise generated by the pump is a sign that the bearings are running dry, but this may not necessarily be accompanied by a fall-off in the pumping rate.

The Dymax pump runs hot; always allow the body of the pump to cool for several minutes before attempting to touch it, since it may reach 70^oC in a normal room. A pump at room temperature, particularly one which has been stored for some time, may take several minutes to reach maximum speed, but under normal circumstances this time is not critical.

Note that the volume passed may be reduced if the smoke stain is very dark, owing to blockage of the pores of the filter paper and the resulting increase in resistance to the flow of air. When this resistance rises to produce a pressure drop of 130 mm Hg across the paper, the pump cannot draw air through the system.

ISO 9835, 1993 specifies that the sample flowrate shall be 2 m³ per day, + 0.2 m³ per day , that is 1.80 to 2.20 m³ per day. Data must be rejected if the sample flowrate is substantially outside the acceptable range. It is therefore important that operators check their sample flowrates every time the site is visited , and take action if the flowrate rises above, or falls below, the acceptable range. The daily sampling flowrate (f) is calculated as follows:

eg:

7th April, time 1030, reading R₁ = 52.47 m³
14th April, time 1030, reading R₂ = 66.89 m³
d = 7 whole days

f =	( 66.89 - 52.47)	= 2.06 m3
	7

(ii) For two meter readings R₁ and R₂ , d days + h hours apart :

10th February , time 0920, reading R₁ = 22.08 m³
17th February, time 1140, reading R₂ = 35.59 m³
d = 7 whole days, h = 2 whole hours, m = 20 minutes.

f =	(35.59 - 22.08)	= 1.90 m3
	(7 + (2 +(20/60))/24)

These examples use metric meter readings, but the procedure is the same for meter readings in cubic feet. AEA Technology Environment’s data processing software calculates sample flowrates using times to the minute, but for the purposes of checking flowrates it is usually sufficient to work to the nearest hour or half hour.

If the calculated flowrate is apparently high (that is, greater than 2.20 m³ per day), the operator will need to address the following questions.

If the flowrate according to the meter is checked and found to be low, (that is, less than 1.80 m³ per day), the operator should address the following questions.

The manufacturer, Charles Austen Pumps Ltd, recommends that the whole filter and critical orifice assembly should be replaced every two years. If the flowrate is checked each week, any persistent downward trends in flowrate should be spotted before the flowrate falls below the lower limit of 1.80 m³ per day, giving ample time for a new filter and orifice assembly to be fitted. We recommend that you keep a new filter and critical orifice assembly to hand . This way, replacement can be carried out quickly and easily, without losing any data. They are available from suppliers of Capex pumps (see our list of suppliers), are not expensive, and are supplied ready assembled, with the filter case sealed and the correct length of reinforced tubing on the end. Do not fit a new critical orifice without also replacing the filter, or the new orifice will block very quickly.