Outdoor in Situ Monitoring of Volatile Emissions From Wastewater

odour insitu
of 4
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Sensors and Actuators B 106 (2005) 36–39 Outdoor in situ monitoring of volatile emissions from wastewatertreatment plants with two portable technologies of electronic noses A. Nake ∗ , B. Dubreuil, C. Raynaud, T. Talou  Agro-Industrial Chemistry Laboratory, ENSIACET, 118 route de Narbonne, F-31077 Toulouse Cedex 4, France Available online 17 August 2004 Abstract This paper reports the test of portable electronic noses for outdoor air monitoring of sewage odours directly in the field. Two commercialdevices with two different gas sensor technologies were tested: conducting polymer (CP) and metal oxide semiconductor (MOS) in orderto select the most appropriate one for this application. As a means to improve the classification ability, the influence of environmentalparameters in different data pre-processing algorithms was studied. The algorithm that permitted the least humidity, temperature and daycorrelation as judged with partial least squares (PLS) was selected. The effectiveness of the systems to discriminate between samples wasobserved using principal component analysis (PCA). The results indicated that CP sensors appear to be unacceptable for this applicationwhile MOS sensors tested were better for discriminating between the different odours. Correlation between the sensor responses andCH 3 SH concentration in water shows that location of samples for MOS sensors depends on the water quality.© 2004 Elsevier B.V. All rights reserved. Keywords:  Sensor; Sewage; Environment; Pollutants; Odours 1. Introduction Atmospheric dispersion of malodours produced fromwastewater treatment processes can cause a nuisance tothe adjacent populations. The nature and concentration of volatile organic compounds (VOCs) in wastewater variesfrom one situation to another depending on several factorssuch as water quality or weather conditions. In order tomonitor pollutants rapidly, a real-time monitoring systemneeds to be used. Previous studies have shown that an elec-tronic nose could be capable of detecting and recognisingdifferent environmental odours [1–5], but very few studies are carried out directly in the field [6,7]. The aim of our research is to determine the ability of portable electronicnoses for in situ monitoring of odours from a wastewatertreatment plant under uncontrolled meteorological condi-tions. Two commercial devices with two different tech-nologies (CP, MOS) of gas sensor array were tested. Thedependence on humidity, temperature and day of analysiswas studied for each electronic nose. ∗ Corresponding author. Tel.:  + 33 562 88 57 38;fax:  + 33 562 88 57 30.  E-mail addresses: (A. Nake), (T. Talou). 2. Materials and methods 2.1. Electronic noses Twodifferentportablecommercialinstrumentswereused:a Pen-2 (WMA Airsense Analysentechnik, 10 MOS) anda Cyranose 320 (Cyrano Sciences, 32 CP). For perform-ing the experiments in the same meteorological conditionsthe analyses were carried out at the same time with thetwo electronic noses. The electronic noses operated by a se-ries of cycles alternating 50s of reference air and 50s forthe Pen-2 and 60s for the Cyranose 320 of ambient air di-rectly pumped to the sensors chambers. Due to the highflow of the pumps (400ml/min for the Pen-2) and in or-der to simplify experimental conditions for field monitor-ing, reference air was filtered air (using a charcoal filter)which was sampled in a Tedlar bag from the ambient airof an office located in the wastewater treatment plant. Aportable PC saved raw data and a 12V battery powered allinstruments. 2.2. Landfill site Experiments were performed from five different outdoorlocations of the wastewater treatment plant, four of them 0925-4005/$ – see front matter © 2004 Elsevier B.V. All rights reserved.doi:10.1016/j.snb.2004.05.034   A. Nake et al./Sensors and Actuators B 106 (2005) 36–39  37Table 1Locations of the measured points in the wastewater treatment plantPoint Locationsa Outdoor sludge/bark mixerb Outdoor dedorization towerc A little apart from the treatment worksd Outdoor sludge dewateringe Clarifier near wastewater treatment works and the last one a littleapart from the treatment works (Table 1).These experiments were conducted during 8 days withdifferent meteorological conditions either sunny, cloudy or Table 2Correlation (underlined) and validation coefficients results for the Cyranose 320 and the Pen-2 for the significant sensors ( P  <  0.05) with the humidity,temperature and day of analysis using various pre-processing algorithmsAlgorithm Cyranose 320 (32 CP) Pen-2 (10 MOS)Number of significantsensorsH  T   Day Number of significantsensorsH  T   Day1/   R  25 0 . 94 0 . 91 0 . 99 5 0 . 67 0 . 81 0 . 610.89 0.88 0.99 0.64 0.79 0.591-  R norm  17 0 . 87 0 . 88 0 . 99 10 0 . 83 0 . 82 0 . 750.84 0.86 0.99 0.77 0.76 0.68   R  /   R 0  0 9 0 . 83 0 . 83 0 . 780.81 0.80 0.72Log  R  25 0 . 95 0 . 90 0 . 99 5 0 . 68 0 . 82 0 . 610.88 0.88 0.99 0.64 0.81 0.58Log  R  /log  R norm  17 0 . 88 0 . 90 0 . 99 10 0 . 89 0 . 86 0 . 840.83 0.87 0.99 0.87 0.82 0.79Fig. 1. PCA between samples based on 17 sensors responses (1-  R norm ) with the Cyranose 320. rainy days. The experiments were carried out at the sametime with the Cyranose 320 and the Pen-2 (120 observationsfor the Cyranose 320 and 123 observations for the Pen-2).Various meteorological parameters were measured: temper-ature, humidity, wind speed and wind direction. Also, odourintensity and odour character, as perceived by the operator,were noted. 3. Results and discussion Five different algorithms were studied in order to selectthe most appropriate one for our study. Only significant sen-  38  A. Nake et al./Sensors and Actuators B 106 (2005) 36–39 Fig. 2. PCA between samples based on five sensors responses (1/   R ) with the Pen-2. sors in discriminating between samples showed by ANOVAresults were used for data processing. Partial least squares(PLS) was used to find the algorithm that has the minimumcorrelation with the humidity, temperature and day of anal-ysis (Table 2).High correlation coefficients were obtained with the Cyra-nose 320. This shows that CP sensors are more dependenton humidity, temperature and day of analysis than MOS sen-sors. The algorithms selected were the complement of thenormalised resistance for the Cyranose 320 and the inverseof the resistance for the Pen-2.Figs. 1 and 2 show principal component analysis (PCA)applied to the data pre-processing algorithm that has theminimumcorrelationwiththehumidity,temperatureanddayfor each electronic nose. Numbers after the point label arerelated to the day of analysis (from 1 to 8).The PCA based on CP sensor results (Fig. 1) showed thatsamples are located depending on the day of analysis: thedrift of the sensors seems to be more important than thesensitivity for this application. Canonical statistics showedthat only the first principal component was significant.MOS sensors seem to be more effective in distinguishingthe samples (Fig. 2). Location of samples depends mainly on the water quality and odour concentration both related byfactor 1: the intensity assessed by the operator and CH 3 SHconcentration in water on day 8 were remarkably higherthan the other days. Also, quite good correlation coefficient(0.79) between electronic nose data and CH 3 SH concen-tration in water was found. This relationship shows thatchanges in sensor responses can be correlated to changes inwater quality.Location of samples appears to depend also in odour char-acter: three mainly clusters (d samples, a–b samples and c–esamples) can be observed even if some samples are out of their group. This mixture of samples can be explained by thefact that odour character and odour concentration in wastew-ater treatment plants varies from one day from another. Thisvariability is due to several factors like water quality and/orweather conditions. These results demonstrated that MOSsensors tested were sensitive to volatile emissions from awastewater treatment plant. 4. Conclusion This experimental study shows that EN technology basedon MOS sensor tested appears to be suitable to outdoormonitoring. On the other hand, CP sensors appear to beunacceptable for this application. In order to enhance themeasurement selectivity and sensitivity further work willfocus on the utilisation of a preconcentrating step before ENanalysis. A prototype will be developed for this application.Also data processing will be optimized in case of continuousmonitoring. Acknowledgements The authors wish to gratefully acknowledge financial sup-port for this work by Générale des Eaux and Midi-PyréneesRegional Council.   A. Nake et al./Sensors and Actuators B 106 (2005) 36–39  39 References [1] P.J. Hoobs, T.H. Misselbrook, B.F. Pain, Assessment of odours fromlivestock wastes by photoionization detector, an electronic nose, olfac-tometry and gas chromatography–mass spectrometry, J. Agric. Eng.Res. 60 (1995) 137–144.[2] T.H. Misselbrook, P.J. Hoobs, K.C. Persaud, Use of an electronicnose to measure odour concentration, J. Agric. Eng. Res. 66 (1997)213–220.[3] K.C. Persaud, S.M. Khaffaf, R.W. Sneath, P.J. Hobbs, Assessment of conducting polymer odour sensors for agricultural malodour measure-ments, Chem. Senses 21 (1996) 495–505.[4] R.M. Stuetz, R.A. Fenner, G. Engin, Assessment of odours fromsewage treatments works by an electronic nose, H25 analysis andolfactometry, Wat. Res. 33 (1999) 452–461.[5] A.C. Romain, J. Nicolas, V. Wiertz, J. Maternova, Ph. André, Use of a simple tin oxide sensor array to identify five malodour collected inthe field, Sens. Actuators B Chem. 62 (2000) 73–79.[6] J. Nicolas, A.-C. Romain, V. Wiertz, J. Maternova, Ph. André, Usingthe classification model of an electronic nose to assign unknownmalodors to environmental sources and to monitor them continuously,Sens. Actuators B Chem. 69 (2000) 366–371.[7] W. Bourgeois, G. Gardey, M. Servieres, R.M. Stuetz, A chemicalsensor array based system for protecting wastewater treatment plants,Sens. Actuators B Chem. 91 (2003) 109–116. Biographies  Almudena Nake  is currently preparing her PhD thesis, sponsored byGénérale des Eaux, at the Agro-Industrial Chemistry Laboratory at theNational School of Chemistry of Toulouse (INPT-ENSIACET). She ob-tained her Master in Environmental Sciences in 2000 at the IndustrialOrganisation School of Madrid (EOI).  Brigitte Dubreuil  is a chemical research engineer. She completed her PhDdegree in 1998, from INPT. Her research interest is focused on odouranalysis with electronic nose using different technologies of gas sensorsand in various fields. Christine Raynaud   is a chemical research engineer in the Aromas andSensory Metrology Group of the Agro-Industrial Chemistry Laboratory.She completed her PhD degree in Analytical Chemistry in 1993 and herHDR degree (research supervising diploma) in 2002, all from INPT. Herresearch interest is focused on technical and sensory analysis. Thierry Talou  is a chemical research engineer and the scientific man-ager of the Aromas and Sensory Metrology Group of the Agro-IndustrialChemistry Laboratory. He completed his DEA in 1984 and his PhD de-grees in Aroma Chemistry in 1992, all from INPT. His research interestis focused on flavour analysis both with classical techniques and srcinalones, like electronic noses.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks