Display PDF in separate tab
Contents
1. neg 4 T 9 er Venter RA Ges NRANeval 03 1N93 14 c 6 9 E wW 2 m 3 1 PN ay V 9 2 5 9 o 5 i 3 77 9 o o 5 5 H P 1 7 pm edb 204510 Cae 2192 100 0 gt f 9 O L 2 a R 59 C 19 7 2 R gt m 3 1 1 1 9 T DuIosas 10 5 HO 097 309 60 Water WEB a5ub22j 410 15 NRA eval 03 1 93 Fouling and drift intermittent immersion The results of this test are presented in the following graphs rather than in a table because of the quantity of data The results given are the differences in reading between the instrument and Laboratory analysis of a sample taken within 3 minutes of the reading This was done 2 or 3 times each day Up to 20 readings were made with the instrument each day See appendix A for details of the calculations involved and an estimate of uncertainties The author should be contacted if numerical data is required All the probes were stable throughout the test but as previously stated the response of the ammonium probe slowed considerably during the test The drift in the dissolved oxygen sat data follows changes in atmospheric pressure during the test This highlights the necessity of daily calib2. 15 10 04 9000000 Table 2 Water Flow rate m s 04 10 17 20 26 32 35 35 30 24 15 10 04 NRA eval 03 1 93 D O 1 1 1 2 2 1 1 1 1 1 2 2 2 22 2 Instrument Ammonium Ammonia ng l mg l 2 5 0 27 3 0 0 26 3 2 0 25 2 2 0 25 3 3 0 27 3 4 0 28 3 4 0 30 3 3 0 25 3 2 0 23 3 3 0 24 3 2 0 24 3 1 0 25 3 0 0 29 2 9 0 28 2 6 0 31 12 mn ON Reading difference Grant Reference Turb Temp c 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KHON Nee NIU 26 36 36 39 39 43 31 38 81 63 64 63 54 38 10 pH Instrument reading difference different water flow rates 31 9 reading difference at different water flow rates Reading difference Grant Reference Total ammoniacal Nitrogen mg l 1 2 2 2 2 2 2 2 52 2 2 2 2 2 s We A P C mo ceo Q m to Uu u Nitrogen Arroniaca Te birdi ty Total Lt 4 V v Li R 5 Bag 27 39960 1 0 891 3192 3109 6 1110 14 691 14 94 83884110 gt M C gt 2 4 gt x gt O H 3 2 m 0 2 gt
3. THE EVALUATION OF GRANT YSI 3800 WATER QUALITY MONITOR 0424 03 ENT AGENCY NRA Instrument Evaluation Centre j Evaluation Fobney Mead Rose Kiln Lane Centre Reading RG2 OSF THE EVALUATION OF A GRANT YSI 3800 VATER QUALITY MONITOR A J Chappell BSc Contents Executive Summary l Introduction 2 Major findings 2 1 Implications for use 2 2 Instrument performance 2 3 Comments on use construction and documentation 2 4 Comments from field use 2 5 Manufacturer s comments E 3 Details of instrument evaluated 4 Test Results 5 Test procedures 6 Definitions abbreviations and references Acknowledgements Appendix Calculations and reference measurements 2 NRA eval 03 1 93 1 Executive Summary This report describes the evaluation of a Grant YSI 3800 multi parameter water _ quality meter is designed for field use and measures dissolved oxygen temperature pH turbidity conductivity and ammoniun The instrument responded well different flow rates and the ammonium probe was the only sensor to exhibit any response change with flow Water temperature affected the accuracy of all the sensors except turbidity and temperature The estuary results imply that the instrument could be used to support or replace the oxygen and chloride titration carried out during a regular monitoring run The ammonium probe response time changed considerably during
4. There was no significant drift of any determinant but the time needed to measure ammonium ammonia increased from a few minutes to approximately 30 minutes which is not suitable for field use The details for each sensor are given below The dissolved oxygen probe did not drift significantly compared to regular Winkler tests A drift of 4 on the 4 saturation range show by the instrument is probably due to a change in the atmospheric pressure during the test this would not cause a problem during normal operation if it is calibrated each day The average difference error and standard deviation random error compared to winkler test is 0 46 0 26 mg l The temperature was stable within measurement uncertainty during the test The average difference error and standard deviation random error compared to a type E thermocouple is 40 09 0 05 C The pH difference compared to laboratory values fluctuated by less than 0 2 but showed no drift during the test The average difference error and standard deviation random error compared to laboratory analysis is 0 007 0 07 During the test the conductivity difference compared to laboratory values fluctuated from 51 5 to 10 uS cm with no apparent drift during the test The average difference error and standard deviation random error compared to laboratory analysis is 414 7 13 1 pS cm The turbidity difference compared to on Line measurements fluctuated between 0 and 5 NT
5. given by A A TotNx 17 14 Hence the levels of ammonium and ammonia in the original sample are calculated from a 17x10 NH SAS tie IN 18 x10P and 17 ig NH A gt wi 12 207 Total Ammoniacal Nitrogen The laboratory measurements were made using an automated flow injection system with a traceable uncertainty of 20 of reading The instrument values were calculated from the ammonium and ammonia readings using the following equation TotN NH 14 gt ACTA NH x NRA eval 03 1 93 25
6. the drift test increasing to approximately 30 minutes to obtain a stable reading The calibration of all the probes was stable throughout the test The storage cup for the sonde does not reflect the quality and price of the instrument and is difficult to remove especially when wet On a number of occasion the instrument software locked up and could only be reset by opening the logger and switching a dip switch on the main circuit board The instruction manual is poorly organised and difficult to use Grant YSI are still developing some aspects of the instrument particularly the ammonium probe NRA eval 03 1 93 3 1 Introduction This report describes the tests carried out to establish the reliability of data produced by a Grant YSI 3800 Tests have been carried out on a single instrument taken new from those supplied to Thames Region for field use The Grant YSI 3800 meter is a multi parameter instrument used for field monitoring of water quality It has six sensors to measure temperature dissolved oxygen pH conductivity turbidity and ammonium From this data values for dissolved oxygen mg l ammonia and salinity are derived The tests completed and reported here studied the effect of water flow rate water temperature water salinity and instrument drift in one month The tests were carried out on all instrument parameters in natural water taken from the river Kennet at the NRA National Evaluation Centre at Fobne
7. 1 91 per The traceable uncertainty of the laboratory values is 20 of reading Salinity The salinity values of sea and estuarine water were calculated from laboratory measurements of chloride in samples submitted for analysis by the following relationship SALINITY CHLORIDE x 1 80655 Turbidity Turbidity was measured in the laboratory using a Hach 2000 turbidimeter and these results were used to check the values obtained in the temperature and flow tests For the drift test the turbidity readings were compared against a pHox 750L on line turbidimeter NRA eval 03 1 93 24 Ammonium and Ammonia An equilibrium is set up between ammonium and ammonia gas in an aqueozs solution by a reversible reaction involving hydrogen ions NH C NH HO The relationship between ratios of ammonium and ammonia and pH are known so the instrument measures ammonium directly using an ISE then calculates the ammonia concentration from the ammonium level pH and temperature The reference values were calculated from the laboratory measurements of Tota ammoniacal Nitrogen TotN and pH pH and the site measurements of absolute temperature in Kelvin T The calculations are based on the following relationship between ammonium and ammonia in water 10 8 NH 105 where c 2729 89 10 1105 0 000071T The flow injection system adds alkali to the sample before measurement and tie level of ammonia in the sample after that is
8. Os ut 5 U 0 gt 2 O 5 f 3 5 out TOR MES VON gt 195 001092 HO 491 109 6 49 1444 3290323j 1D Deo C gifference grant Water temperature Tables 3 and 4 below and the following figures show che difference between the instrument output and the analytical results at different water temperatures The test was carried out in an isolated water sample in a tank at the evaluation centre See appendix A for detaiis of the calculations involved and an estimate of uncertainties Table 3 Instrument reading difference at different water temperatures Water Reading difference Grant Reference Temperature D O D O Temp Turb pH C mg l C NTU 4 6 4 9 0 8 0 0 2 3 0 46 9 2 8 2 1 1 0 1 2 2 0 47 19 8 9 3 1 0 0 1 3 4 0 41 30 2 11 7 0 9 0 1 1 0 0 20 40 4 9 8 0 7 0 2 1 3 0 09 45 4 9 3 0 6 0 2 1 3 0 07 45 1 8 6 0 5 0 1 3 0 0 14 19 8 5 5 0 6 0 0 3 0 0 53 Table 4 Instrument reading difference at different water temperatures Water Reading difference Grant Reference Temperature Ammonium Amnon a Total ammoniacal N Cond C mg l ng l mg l amp 6 1 3 0 23 0 8 7 8 9 2 2 8 0 28 1 9 5 8 19 8 3 0 0 37 2 0 91 30 2 1 5 0 26 0 9 16 7 40 4 0 5 0 19 0 2 28 6 45 4 0 3 0 26 0 0 82 5 45 1 1 8 0 12 1 3 21 1 19 8 3 0 0 74 1 7 9 1 Temperature omauct tivity Corrected to 25 deg C 5 mie
9. U but showed no drift during the test The average difference error and standard deviation random error compared to on line measurements is 1 2 1 0 NTU The variation of total ammoniacal nitrogen measurements compared to laboratory values indicate the variations in the ammonium electrode combined with the calculations of ammonia within the instrument The total ammoniacal nitrogen differences fluctuated between 1 5 mg l and 2 7 mg l and show no significant drift during the test The water was dosed to approximately 8 mg l by addition of ammonium chloride The average difference error and standard deviation random error compared to laboratory analysis is 0 02 0 98 mg 1 HOWEVER these readings were taken after the instrument reading had settled and so hides the significant change of response time which occurred for the ammonium measurements At the start of the test the reading settled within 5 minutes but near the end of the test it took approximately 30 minutes to settle which would not be practical for field use This was discussed with Grant who suggested fouling of the electrode A thick grey film had developed over the membrane and the settling time was reduced to approximately 15 minutes when this was removed using de ionised water from a wash bottle At the end of the test the response had increased again to approximately 20 minutes Since the settling time was not measured during che test these values are estimates NRA e
10. as measured directly with the sonde In this case the meter should read 100 when the water is at the calibration temperature The highest value reported was 95 saturation Could some of the error be due to inadequate saturation and equilibration of the water as temperature was varied We believe that the variation associated with temperature and drift with time would be eliminated by carrying out the recommended calibration immediately before making measurement The DO drift observed seems reasonable but it is unlikely that all 4 drift would be due to atmospheric pressure alone ii pH Sensor The offset between the instrument and laboratory measurement with changes in temperature could be due to the instrument being set for automatic buffer temperature correction during calibration which is incorrect if non Grant YSI pH calibration solutions are used This would cause an estimated error of approximately 0 2pH It should also be noted that the temperatures are not compensated to 25 C as indicated in our manual and the NRA report Our manual is in error on this point and will be corrected iii Turbidity Sensor The scattered measurements with changes in water flow rate are almost certainly due to air bubbles from the air saturated water in the flow tank forming on the sensor The bubbles would not be present in still water or water taken to the laboratory because the sample would de aerate in transit iv Ammonium Sensor The problem
11. em there were no corrections made for salinity in either the reference measurements or the 3800 measurements Because of this it is not possible to attribute the trend show n these results to changes in the instrument The conductivity reference values have an uncertainty of t3 mS cm Ammonium measurements are not included because the electrode response time was too long and because of the large interference caused by sodium and potassium in the sea water The turbidity results are omitted because of uncertainty in the reference measurements The temperature probe was not tested Table 5 Instrument reading difference at different salinities Water Reading difference Grant Reference Salinity Salinity Cond D O D O pH to to mS cm mg l 37 6 3 2 0 8 _ 0 9 0 2 0 6 37 6 3 2 0 4 0 8 0 4 0 6 36 7 3 2 1 7 1 6 0 4 0 5 34 3 3 3 0 8 0 8 0 4 0 4 29 3 0 5 1 3 3 2 0 4 0 4 26 7 2 2 0 3 2 1 0 4 0 3 20 8 0 7 2 9 2 2 0 3 0 2 15 5 0 2 2 3 0 3 0 2 12 0 0 3 2 6 0 3 0 1 9 0 0 1 3 2 0 4 9 0 0 3 0 0 6 1 1 4 4 0 2 6 0 3 0 1 5 6 0 0 0 2 2 1 0 3 0 1 3 4 0 2 1 4 1 0 0 2 0 0 2 4 0 1 1 3 0 2 0 1 1 4 0 3 0 7 0 9 0 2 0 2 NRA eval 03 1 93 21 lt 5 Test procedures Water Flow Rate The probe was calibrated according to manufacturer s instructions ie dissoived oxygen was calibrated at 100 turbidity 0 200 and 800 FTU pH 4 7 and 10 and ammonium at 1 and 100 mg l The additional temperature and conductivity cal
12. ements are made by Winkler titration The uncertainty is assumed to be 0 5mg l from observation For sat values the water is aerated to 1008 to other values by bubbling air or Oxygen Nitrogen certified mixtures through river water The uncertainty is assumed to be 3 from observation Saline Dissolved Oxygen For mg l values the reference measurements are made by Winkler titration The uncertainty is assumed to be 0 5mg l from observation The sat values are obtained by calculating the oxygen solubility from the temperature and the chloride values using the tables given in BS6068 then calculating the saturation from the vinkler values The chloride levels were obtained by laboratory analysis of the samples submitted Temperature The reference measurements were made using a type E thermocouple with the cold junction compensation within a Schlumberger IMP data acquisition unit The uncertainty is taken as t0 1 C pH The pH is compared against laboratory measurements made using an automated CSP pH meter During the drift test the readings were confirmed by comparison with the monitor panel The traceable uncertainty is 5208 of reading Conductivity The conductivity is compared against laboratory measurements made using an automated CSP conductivity meter The laboratory readings were given as conductivity at 20 C this adjusted to 25 C for comparison with the instrument values using a temperature coefficient for the water of
13. he ammonium probe is useful for special surveys where it is continuously immersed but quickly becomes useless for routine sampling tests because of the response time The results from the flow test show that the sonde does not need to be kept moving to obtain an accurate DO reading Also the variations caused by temperature between 5 and 25 C are smaller than the values above The conductivity and pH readings are both corrected to a temperature of 25 C optional for the conductivity The Laboratory readings are corrected to 20 C and this difference may cause some confusion and can generate differences of 11 of reading in conductivity values The values obtained for D O pH and salinity during a normal monitoring trip into the Thames estuary were within the measurement uncertainty of the laboratory and the uncertainty of titration carried out on the boat Since the instrument is easy to use it could reasonably replace or support the titration measurements of oxygen and chloride When the problems associated with the ammonium probe are solved then regular calibration and servicing using traceable and approved techniques could lead to accurate and reliable measurements using this instrument 2 2 Instrument performance Water Flow rate Only the ammonium probe showed any significant variation with flow the other probes performed well in this test especially the dissolved oxygen which was significant better than most non stirred DO probe
14. ibration points for ammonium were not set Temperature and conductivity were not calibrated since they are factory set The sonde was then be placed in the large flow tank containing an isolated sample of river water The flow rate was varied from 0 04 to 0 37 m s measured by an electromagnetic flow meter Ammonium Chloride was added to the tank to give approximately 10 mg l ammonium and it was aerated throughout the test to maintain the DO level at 100 The temperature was monitored by a Type E thermocouple and samples were taken during the test for analysis by a NAMAS accredited laboratory of pH conductivity total ammoniacal N and turbidity Water temperature The dissolved oxygen probe was calibrated in air according to the manufacturer s instructions and the amsonium probe was calibrated at 1 and 100 mg l The optional ammonium calibration points to correct for temperature and conductivity were not set The sonde was placed tank of recirculated river water to which was added Ammonium Chloride to give approximately 10 mg l ammonium and which was aerated throughout the test to maintain the DO level at 100 The water temperature was stabilised at the following temperatures and the instrument outputs noted 4 6 C 9 2 C 19 8 30 2 40 4 C and 45 4 C test was repeated at two decreasing temperatures The temperature was monitored by a Type E thermocouple and samples were taken each temperature for ana
15. is not highiy regarded but it is the most demanded sensor as it is an important environmental substance It has been used successfully NRA eval 03 1 93 for in number of surveys For survey work there is a requirement for more versatile logging facilities particularly for a delay of start facility or some form of timer control start and stop logging There have been a number of comments about the batteries These mainly concern the need to open the logger to change batteries Rechargeable batteries are also requested 2 5 Manufacturer s comments Grant Instruments provided the following comments on the report We have separated out sensor performance and indicated our understandings based on the findings of this report and our own observations It was however not possible to interpret all the results without detailed knowledge of the experimental procedure It should be noted that tests were carried out on a single logger and sonde and the results cannot be viewed as indicative or definitive for the product line as a whole Overall we believe that the report is thorough in its approach and will turn out to be reasonably favourable with regard to the 3800 performance A Sensor Performance i Dissolved Sensor We do not fully understand how the temperature dependence of the DO probe accuracy was assessed We would assume that the instrument was calibrated in air at 100 and that water sparged with air w
16. lysis by a NAMAS accredited laboratory of pH conductivity total ammoniacal N and turbidity Fouling and Drift intermittent immersion The probe was calibrated according to manufacturer s instructions ie dissolved oxygen was calibrated at 100 turbidity 0 200 and 800 pH at 4 7 and 10 and ammonium at 1 and 100 mg l The additional conductivity calibration points for ammonium were not set Temperature and conductivity were not calibrated since they are factory sec The sonde was then regularly immersed up to 20 times in class 1 river water then returned to its storage condition on each working day for one month In order to generate a measurable level of ammonium it was necessary to trap a sample of water and add a known amount of ammonium chloride The temperature was monitored by a Type E thermocouple and samples were taken during the test for analysis by a NAMAS accredited laboratory of pH conductivity total ammoniacal N turbidity and dissolved oxygen The trend in levels of D O cemperature pH conductivity ammonium concentration and turbidity of the water were monitored using standard water quality instruments The sonde was also occasionally immersed in the fish breeding ponds at Fobney NRA eval 03 1 93 22 be ON Mead and was used once in estuarine conditions at Crossness During che last week of the test the sonde was placed in the intermittent immersion tank Fob
17. ney to be immersed automatically up to 10 times because of the pressure of time Salinity The instrument was taken on a normal monitoring run into the Thames estuary It was mounted in a bucket with a continuous supply of river sea water taker from the normal monitoring supply pumped into the boat Measurements were made in water with salinity levels varying from 3680 to 140 and were compared laboratory analysis of samples taken at the same time Winkler titration were carried out on the boat as part of the regular monitoring along with chloride titration Laboratory analysis provided values for pH conductivity chloride ammonium concentration The reference salinity values were calculated fron laboratory chloride analysis 6 References A Chappell and Loewenthal Test protocol for Grant YSI 3800 water quality meter Loewenthal Evaluation report on Grant 3800 water quality meter I Baldwin Protocol for investigation of ion selective ammonium electrodes and their applications in field measurement BST BS6068 section 2 15 1986 Determination of dissolved oxygen electrochemical probe method D Neville et al NRA evaluation reports on DO meters Acknowledgements Thanks must be given to Mr M Loewenthal and the AWQMS staff and to the Laboratory staff at Fobney Mead NRA eval 03 1 93 23 Appendix Calculations and Reference measurements Dissolved Oxygen For mg l values the reference measur
18. of increased response time in use is being investigated and NRA eval 03 1 93 9 an alternative sensor is currently on laboratory and field trials and to date no increase in response time has been noted We also wait for results of NRA trails on the new probes In both cases YSI recommend storage of the ammonium probe confidence solution The observed variations in ammonium recording with flow have been observed but not to the extent indicated in the report It may well be that the error is due to slow response observed on this probe v Conductivity Sensor Salinity we were glad that these tests were made and that the results indicate the instrument is useful in brackish water B Other Points i Lock up We do not understand the reported lock up and we have not received other reports with version 4 firmware which is current in all NRA 3800 loggers We would like you to make further tests on a different logger to see if the symptoms persist ii Screw Sonde Cap We agree entirely with your point on the plastic transit cap A screw on rigid cap has been designed and will be available from the end of February and can be ordered for retro fitting at 82 less NRA discount iii Battery Access Two alternative proposals to overcome the need to open logger case are being investigated with the view to designing one which can be ordered for retrofitting iv Batteries YSI are investigating use of alkaline rechargeable ba
19. r periods of time prior to recording and reading making sure that che sparged wazer reads close to 100 saturation at the air calibration temperature We take the opportunity to thank the NRA for investing time and effort in obtaining a better understanding of water quality measurement in the field It is from this type of basic work that we at Grant YSI can further develop and improve our instrumentation in line with your requirements 3 Details of instrument evaluated The instrument teste was selected at random from a batch of instruments purchased by the NRA and supplied to Thames region for normal use Instrument Logger Serial Number Sonde Serial Nunber NRA reference Number Manufacturer Tel Fax Source NRA eval 03 1 93 Grant YSI 3800 Water Quality Logger 381 200 444 381 500 505 31L 444 5 505 Grant Instruments Cambridge Ltd Barrington Cambridge CB2 502 0763 260811 0763 262410 NRA Thames region 11 4 Test Results Water Flow Rate Tables 1 and 2 below and the following figures show che difference between the instrument output and the analytical results at different water velocity The test was carried out an isolated water sample in a flow tank at the evaluation centre See appendix A for details of the calculations involved and an estimate of uncertainties Table 1 Water Flow rate m s 04 10 17 20 26 32 35 35 30 24
20. ration of the sensor during use to reduce this effect This drift is not apparent when the mg l values are compared to winkler analysis Temperature Deg C T C difference grant days NRA eval 03 1 93 16 40 gt p E M 5 ps S 2 5 20 x jee i M gt e 0 0 U C3 f a ie 25 m 8 paro 1 2 ns a U di Caa xu A 0 82u828jJ1p 125 indino fs E z 0 IPA9NVUN 81 sAep difference grant calc Lab mg CLO lab pH difference Cgrant SAED 2 2 2 LA difterence grant calc Lab catc mg l o r s No Le shen difference grant tab calc mg l sAep Un UOUIWY FTU Panel difference grant NRA eval 03 1 93 p r3 m 20 uS cm lao difference Cgrant 2 correzied 1 days 25 cez V Uy Salinity Table 5 below shows the difference in readings obtained between the grant YSI 3800 and laboratory results The uncertainties quoted below are estimates The salinity values are calculated from laboratory analysis of chloride levels The dissolved oxygen reference values are from winkler titration carried out on the boat The pH reference measurements were made with a Russel electrode within a CSP syst
21. s For use in high flows a sinker weight is necessary because the sonde tends to swing on its cable and lie horizontal on the water surface NRA eval 03 1 93 5 The dissolved oxygen showed a slight increase at flows greater than 0 3 m s but the variation was less than 1 throughout the test which is less than the reference uncertainty This is significantly better than most non stirred DO meters The temperature difference varied by 0 1 C during the test which is similar to the reference uncertainty The pH difference varied by less than 0 1 during the test The conductivity difference varied by less than the measurement uncertainty during the test The turbidity sensor reading increased during the test but this was not related to the water flow The reason is unknown but it could be due to fouling air bubbles forming on the probe surface or cavitation around the sonde head The total ammoniacal nitrogen difference varied significantly from 1 7 mg l at O m s to 2 4 mg l at 0 35 m s The water was doped to approximately 8 6 mg l by addition of ammonium chloride Water temperature The turbidity and temperature probes showed no significant change with temperature however the other probes all changed with temperature The dissolved oxygen 8 reading varied between 95 1 85 at 4 6 C and 88 3 sat at 30 2 C in aerated water There is an apparent reduction in output of 0 14 sat C This is larger than the manufact
22. sed by vibration and may indicate a problem with the board mount or the switches and could result in batteries failing A rigorous vibration test may provide some useful information for the manufacturer about this problem The manual contains a lot of information about the instrument software and available probes Unfortunately it is very badly organised with most of the useful information contained in 2 sets of addenda approximately 50 pages long including addenda to the first addendum As a result the manual is very difficult to use There is no information about cross sensitivities of the sensors this is important since ammonium ion selective electrodes are known to be sensitive to potassium and sodium which would give a false positive reading in sea water an order of magnitude greater than an ammonium response The logger has been designed to IP65 ie dust proof and splash proof but as there 1s a risk that it could be accidentally immersed in a river this may be inadequate IP67 ie dust proof and can vithstand immersion to lm deep would be more appropriate 2 4 Comments from field use This is a summary of work carried out by M Loewenthal which is reported separately and of comments from pollution officers The manual is regarded as poor difficult to use and tends to make use of the instrument seem more complicated than is the case This has made one to one training necessary when issuing instruments The ammonium probe
23. tteries for use in the 3800 We will keep you informed on our findings and would encourage the NRA to test these or similar batteries v Instruction Manual We acknowledge the comment made on the shortcomings of our user manual An updated and revised manual will be produced to overcome the criticisms and we plan to have this available in March 1994 vi Waterproofing The logger is water resistant to 65 proof against low pressure jets of water For proofing against immersion we recommend fitting the logger 66 waterproof case We have recently quoted for 50 such cases after producing a prototype CONCLUSION This report covers a wide range of tests and experiments on the 3800 Water Quality Logger and could obviously be the basis of further investigation The following four experiments are not very time intensive and could be fairly definitive in determining if some of the problems encountered were due to the particular 3800 user a With a different 3800 transfer the sonde from confidence solution to river water spiked with 10mg L ammonium and check the time required for stabilisation on a sonde containing one of the Analytical Sensors ammonium probes supplied recently NRA eval 03 1 93 10 b c Run a seperate logger over a 4 5 week basis to see if any lock ip occurs Re run the effect of temperature change on dissolved oxygen accuracy but sparge the water more vigorously with air and for longe
24. urer s specification of 0 3 between 5 and 45 C After conversion to mg l the reading difference compared to the theoretical value for saturated water varied from 1 13 mg l at 9 2 C to 0 54 mg l at 45 1 C and it increases with temperature by approximately 0 01 mg l per C This is also outside the manufacturer s specification of 0 6 reading between 5 and 45 C The temperature difference decreases slightly with temperature however since this is within the uncertainty of the reference system then it 15 not significant The pH reading difference compared to laboratory values changed from 40 47 at 9 2 C 40 07 at 45 4 C The conductivity difference compared to laboratory values varied from 46 yuS cm at 9 2 C to 442 pS cm at 45 4 C The difference increases with temperature above 20 C The turbidity difference varies between 1 0 NTU at 30 2 C and 3 4 NTU at 19 8 There does not seem to be any significant variation with temperature The total ammoniacal nitrogen difference compared to laboratory values varied from 2 0 mg l at 19 8 C to 0 0 mg l at 45 4 C The water was dosed to approximately 7 5 mg l by addition of ammonium chloride NRA eval 03 1 93 6 Fouling and Drift intermittent immersion The only fouling observed during the test was a grey film which deve loped on the ammonium probe membrane This contributed to the slow electrode response and its removal only partially restored the probe
25. val 03 1 93 7 Salinity The instrument was taken a regular monitoring run to the lower Thames Estuary the differences observed bezveen the instrument readings and the laboratory results are within the laboratory uncertainty From this comparison it is not possible to find anv effects of salinity on the instrument 2 3 Comments on use construct on and documentation The instrument generally robust and vell made but there are a number of things which cause problems in normal use The probes are protected by a tight fitting plastics cup which fits over the outside of the sonde and is sealed by an O ring When a good water tight seal is made it is very difficult to remove the cup more so when wet and the water inside was spilt on a number of occasions This is impracticable and will sometimes make it difficult or impossible to use the instrument on a river bank without help to remove the cup The screw on cup now being developed is preferable On a number of occasions the instrument software locked up and did not respond to any keystrokes This had to be rectified by opening the logger and resetting it using a switch on the main c rcuit board The batteries are also changed by opening the logger Both operations need to be carried out where the circuit board can be protected from damp dust splashes etc When transported by car the instrument switched itself on this has also been observed by pollution officers This could be cau
26. y Mead and in the Thames Estuary The instrument output was compared to analysis by the NAMAS accredited laboratory at Fobney Mead The tests were carried out over four months to the end of November 1993 to a test protocol agreed by Grant YSI and the NRA Further tests are being made on an instrument with a different ammonium probe after recent developments by Grant YSI NRA eval 03 1 93 4 2 Major findings This section provides a summary of the test results and_includes subjective comments about the instrument Comments arising from field use of 30 other instruments in Thames region are also noted here For details of the instrument performance during the test refer to section 4 2 1 Implications for use The results of these tests can be used to give an estimate of the measurement uncertainty of the instrument for sampling measurements If this is taken as three times the standard deviation of the drift test to give 95 confidence values we get the following values Dissolved Oxygen 6 sat typically 1008 Temperature 0 2 typically 15 C pH 0 2 typically pH 8 Conductivity 39 pS em typically 550pS cm Turbidity 3 NTU typically 2 NTU Since the ammonium probe requires a long time to obtain a stable reading and because of the known interference from potassium and sodium then an uncertainty value can not be produced for the ammonium and ammonia values Pollution officers experience shows that t
Download Pdf Manuals
Related Search