Odour Prevention and Control of Organic Sludge and Livestock Farming
Odour Prevention and Control of Organic Sludge and Livestock Farming
Edited ByV.C. Nielsen, J.H. Voorburg
Edition 1st Edition
First Published 1986
eBook Published 29 April 1986
Pub. location London
Imprint CRC Press
Pages 404 pages
eBook ISBN 9780429077852
SubjectsEngineering & Technology, Environment & Agriculture
Nielsen, V. (Ed.), Voorburg, J. (Ed.). (1986). Odour Prevention and Control of Organic Sludge and Livestock Farming. London: CRC Press, https://doi.org/10.1201/9781482286311
Proceedings of a round-table seminar, Silsoe, UK, 15-19 April 1985.
TABLE OF CONTENTS
LIQUID AGRICULTURAL WASTES" A. M. Bruce Water Research Centre, Stevenage, Herts Co-ordination and co-operation among EEC and other countries in the field of research on sewage sludge has been active since 1972 under the
COST 68 bis was more extensive in scope than the initial project since it covered both treatment and use of sewage sludge. This meant that the topics included both the engineering and economics aspects of sludge processing at sewage works and the environmental aspects of sludge disposal particularly in regard to its utilisation as a fertilizer in agriculture. The use of sewage sludge in this way is important in most countries and it was recognised that co-ordinated research was desirable into both the possible adverse environmental effects of heavy metals and pathogens in sludges and the beneficial effects of plant nutrients in sludge. The problem of odour nuisance arising from the handling and spreading of sewage sludge was also recognised as an important subject-area for research. Under COST 68 bis, five Working Parties were established to co-ordinate the various areas of research. As a 'Concerted Action1, no direct funding was available from the European Commission to finance research projects on sewage sludge, each country being expected to contribute its own publicly-funded projects to the common ’pool*. COST 68 bis ran from 1977 to 1980(2) and was followed by an extension programme - COST 68 ter - which ran from 1981 to 1983(3). In 1983, it was decided to further extend this Concerted Action but to widen the range of research topics in the scientific programme to include animal manures. The renewed programme was designated COST Project 681 'Treatment and Use of Organic Sludges and Liquid Agricultural Wastes'. It is obvious that sewage sludges and farm manures have many aspects in common particularly with regard to handling and treatment techniques and to the environmental impact (e.g. odour) which can occur from their utilisation on land. On the other hand, from the administrative point of view, sewage sludge and farm manures are in two different 'worlds'. Sewage sludge is the general responsibility of public authorities while responsibility for disposing of animal manures belong mainly to the private farming sector. Funding for research on the two types of waste, even if from Government sources, is usually from different Departments and there is little cross-involvement of research scientists in the two sectors. Nonetheless, the COST 681 activity is attempting to promote some co-ordination of effort between the two research areas and, hopefully, this will result in mutual benefit to both those authorities responsible for sewage sludge treatment and those concerned with farm manures and their disposal. This joint Workshop on 'odours' is a good example of the type of co-operation, and sharing of information on a common problem, which can be of great mutual benefit to both sectors. PARTICIPATING COUNTRIES
The countries participating in the current COST 681 activity are:- Other European Countries Belgium Austria Denmark Finland France Norway
The non-EEC European countries have to make a financial contribution to the European Commission in respect of the COST 681 activity and this is a sure indication that they consider their involvement to be of value. Canada has the more informal status of an invited participant in sane of the technical meetings. There are also links with relevant international organisations such as the FAO - as is evidenced by this joint Workshop. ORGANISATION QE COST 6S1 ACTIVITIES Working Party 1. Working Party 3. Hygenie Aspects Working Party .4. Agricultural Value
As indicated earlier, COST 681 is organised into 5 Working Parties under the general guidance of a Management Committee of National Delegates (Current President: H. M. Scheltinga of the Netherlands). The Working Parties and their respective areas of responsibilities are:- Sludge Processing Working Party 2; Chemical Pollution
complaints or problems relating to disposal of sewage sludge to land were related to odour nuisance. Table 1. Summary of complaints or problems relating to disposal of sewage sludge to land in the UK (1980 data)(^) Type of problem Per cent occurrence Environmental nuisance (smell) 60 Transport 19 Water Pollution 10 Agricultural 5 Metals 4 Veterinary 1 Planning consent 1 100 It is clear from this that odour nuisance is an important problem and it follows that standard methods of scientifically measuring odours are desirable. ACHYHX Iff QQ2I M l . SUBdGTOUP .r.OBMBS' This sub group was established in 1984 as part of the activity of Working Party 1 but drawing on experts from outside the Working Party. The Chairman is Mr J H Voorburg of the Netherlands and other experts in odours include Dr M Hangartner (QD, Dr J Hartung (D), Dr A Eikun (No) and Mr V C Nielsen (UK). Mr H M Scheltinga (NL) and A M Bruce (UK) are also members of the group. The sub-group is hoping to complete its tasks quickly, the main ones being (a) To develop proposals for a harmonised and standardised odour measurement technique (b) To exchange information about research on odour measurement and control. On (a) good progress has been made in collecting information on the existing guidelines in different countries for sampling and transportation of samples for odour measurement, for dilution techniques and for panel selection etc. The question of acceptable levels of odour intensity is net being considered. All these matters will be discussed at this joint Workshop and it is hoped that clear recommendations will emerge from the experts so that a formal report can be presented for discussion at the COST 681 4th Symposium to be held in Rome in October 1985. In regard to objective (b), one of the major tasks of the sub-group has been to compile an inventory of organisations and scientists in Europe who are actively involved in research on the measurement and/or
control of environmental odours. Additionally, the sub-group is preparing a bibliography of European reports and papers on odour measurement and control which have been published in the last 5 years. This inventory will also be made available at the COST 681 Rome Symposium in October 1985. So far, the enquiries for the inventory of research organisations has indicated a considerable variation in research activity on odours among the European countries (Table 2). In most cases, there is seme Government funding for this research. Overall, it is hoped that this aspect of the sub-group’s work will promote an improved inter-change of information and co-operation between organisations and scientists in this field. Table 2. Preliminary information concerning research on the measurement and control of environmental odours in European countries Country No of Research Government organisations involved Funding Belgium France 5 FR Germany 15 ✓ Netherlands
Other activities related to odours are directed mainly at odour control techniques - particularly methods of sludge and slurry stabilisation. There is continuing research and development relating to anaerobic digestion as a process for odour control; a COST 681 Workshop on new developments in anerobic digestion was held in 1984(5). There are
(3) Commission of the European Communities. COST project 68 ter. Final Reports of the Community COST Concertation Committee. 1983. (4) Department of the Environment/National Water Council Standing Committee on the Disposal of Sewage Sludge. Sewage Sludge Survey 1980 Data. Department of the Environment 1983. (5) Anaerobic digestion of Sewage Sludge and Organic Agricultural Waste. Proceedings of a COST 681 Workshop held in Athens May 1984. (In the Press). (6) EIKUM, A.S., and BERG, N. Odour characterisation and removal of odours from facilities receiving septage. To be published in Review Papers on Sewage Sludge Processing, 1985. Commission of the European Communities, Brussels.
(1) Commission of the European Communities. COST Project 68. Sewage Sludge Processing. Final Report of the Management Committee. 1975. (2) Commission of the European Communities. COST Project 68 bis. Treatment and Use of Sewage Sludge. Final Reports of the Community - COST Concertation Committee.
SAMPLING OF ODOUROUS AIR FOR OLFACTOMETRIC MEASUREMENT J. HARTUNG Institute for Animal Hygiene of the Hannover School of Veterinary Medicine, Blinteweg 17 p , 3000 Hannover 71, FRG Summary Both static and dynamic sampling procedures are used for olfactometric measurements. Care must be taken inorder to obtain a representative sample and to minimize sample losses due to condensation, adsorption and permeation, when using static sampling methods, particularly. Teflon or Tedlar bags and inert tubing materials help to diminish adsorption and desorption problems. Condensation can be avoided by heating the sampling tubes or by prediluting the sample with pure, odour-free air. Within the EEC guide lines exist for odour measurement in The Netherlands, France, Germany and the United Kingdom. The usefulness of dynamic sampling is agreed on. The opinions differ as far as static sampling is concerned. It seems that both sam pling methods can be applied successfully for olfactomet-ric measurements. However, it is necessary to define the details of the procedures aiming at a standardization of sampling which might be the first step for a harmoni zation of olfactometric measurements in the laboratories of the different countries. 1 . INTRODUCTION The method of measuring odour sensorily in general can be devided into the following basic steps (1): - sample collection - sample dilution and presentation - indication of response - interpretation of response Due to the fact that many different testing procedures exist in the different laboratories, results can only be com pared when knowing exactly - the conditions and procedures for sampling of the air to be i nvesti gated, - the design and function of the olfactonnetric apparatus, and - the physiological and physical status of the panel. The olfactometric apparatus and the panel are in close connection with each other as shown in Table I whereas the sam pling procedure is more or less apart from the apparatus and the panel and affects the olfactometric inlet, only. However, sample collection is the first step and can influence the re sults considerably; thus, valid sampling is the base for valid
measurements. This paper is confined to the different forms of sampling odourous gases for olfactometric measurements and the problems involved. It refers to existing guidelines for olfactometric measurements in the countries of the EEC, as well. 2. TYPES OF SAMPLING Samples of odourous gas may be collected in unconcentrated or concentrated form. Concentrated sampling is usually neces sary when gas chromatography or other chemical analytical meth ods are to be used. Unconcentrated sampling is provided if o-dour threshold concentrations are required (2). Depending on the type of olfactometer used dynamic sam pling or static sampling are provided. The principle of dynam ic sampling is shown in Figure 1. It requires a part-flow of the odourous gas to be continoulsy extracted from the source and subsequently directed to the olfactometer. This sampling method implies that the measurements are carried out close to the source. An advantage of the method is that there is the possibility of controlling a process, directly, and in case of the break-down of the process this can be noticed right away. A disadvantage of the dynamic method is that odour sources that are not readily accessible require a relatively great ef fort in order to install the olfactometer and suitable sam pling pipes which often should be insulated or heated to avoid adsorption or condensation (3). When static sampling is used a partial stream of the o-dourous air is collected in a sampling vessel. Samples are taken from this vessel or bag to dilute the odourous air for the olfactometer using syringes or on-line tubings. When using this method odour measurement with the panel can be carried out at any arbitrary location, if the vessel is a transport able one. An example for static sampling is given in Figure 2. 3. PROBLEMS OF SAMPLING the main problems encountered when sampling odourous air derive from surface effects of the sampling tubes and vessels, namely by - adsorption, - desorption, and - condensation. This depends mainly on the material of the tube, the vessel or the bag (adsorption) or on the nature of the gas, whether it is hot and/or containes a high amount of humidity (condensa tion). On the other hand the sample can be altered by trace components bleeding from the material of the walls of the ves sel or the tube (desorption). The following factors are to be observed for valid static sampli ng. aTTTToTce of_m£teri aj_ For tWe sampling of odourous gases glas vessels, stain less steel tanks (4) and flexible plastic bags (5) were tested. The initial concentrations of the test gases decrease consider ably with storage time in glass and steel vessels. In recent years bags made of Polyethylene(6), Teflon (3) and Tedlar (7), (8) were usually used. Figure 3 shows a graph from SCHUETZLE
et al. (8) indicating the good properties of Tedlar. ROOS et al. (3) point out that when Tedlar bags are used decreases in concentration up to 60 and 70% are observed with aromatic com pounds. These authors prefer Teflon FEP-bags and report on compa rable results between dynamic sampling and static Teflon FEP-bag sampling, as shown in Table II. Assuming that adsorption and desorption effects are mini mized by the choice of material, reactions between compounds in the gas phase can not be excluded, j) ._The_prevejiti o_n £f_condejisati_oji ConcTensatTon can Fe avoTded by the predilution of the sam ple by dry, odour-free air. It is important to know in which ratio the sample is diluted for odour unit or odour threshold determi nati on. ._0ther_f ac ;to£S_o f_ ijmj)ortance
wFen performing vesseT/Fag sampling: - permeability of the vessel/bag (% of losses), (9), - sampling time (moment value, or time integral), - transport and storage time, - sampling volume (sufficient for repeated measurements),
RECOMMENDATIONS FOR SAMPLING AS DESCRIBED BY GUIDELINES Four guidelines exist in the states of the EEC concerning odour: - The Netherlands: Geurnormering - Odour standard (1983) - France: Norme experimentale X43-101 - Experimental standard, 1st draft, (1982) - Germany: VDI-Guideline 3881, part 2, draft (1984) - U.K.: Odour control - a concise guide (Warren Springs, 1980) Only the guideline of Warren Springs, U.K., and the VDI-Guide-line contain detailed descriptions of the sampling procedure. Concerning both dynamic and static sampling there is an agreement to avoid adsorption and condensation By insulating and heating the inlet tubes, and by prediluting the sample with dry odour-free air when bag or vessel sampling is used. Desorp tion problems can be minimized by preflushing. Inert materials have to be used, only. The storage time of the sample in bags should be as short as possible. If required dust filters can be provided to prevent a contamination of the olfactometer by dust. The most detailed description of sampling is given in the German VDI-Guideline 3881. The recommendations do not prefer a certain technique but accept different techniques if they meet the requirements to prevent condensation, adsorption, influ ences of the dust, and desorption of odorous compounds from the sampling system. The guideline points out that these im portant problems should, if necessary, be settled in prelimi nary experiments. The guideline includes the description of a dilution equipment to avoid condensation. Figure 4 shows the forms of static sampling given in the Guideline. The Warren Springs, U.K. guide gives a full description of the recommended sampling procedure and apparatus, as shown in Figure 5. Tedlar bags of a capacity of 40-50 1 are used.The bags are introduced in a 60 1 carboy. All tubes except for the underpressure tube which is made of metal consist of PTFE in cluding the valves. A small battery-operated pump is used to evacuate the container, which allows the initially collapsed Tedlar-bag to inflate (14). 5. FINAL REMARKS Both dynamic and static sampling procedures are suitable for taking samples for olfactometric measurements (15), (16). If the olfactometer and the panel are available close to the source dynamic sampling may be preferred. The equipment for preventing condensation in the sampling pipe and contamination of the sampling pipe and the olfactometer by dust should be provided. Static sampling may be used at sources of odour that are not readily accessible or where the odour concentrations are changing quickly or because of expenses. When using static sampling the most important requirements are to avoid losses of sample-born compounds by adsorption and condensation and the contamination of the sample by impurities desorbing from the sampling and storing divice. Interaction of the compounds in the sample during storage can be minimized by keeping the time
of storage as short as possible, only; 24 h should not be ex ceeded. Table III comprises the most important criteria for valid static and dynamic sampling. It seems that both the guide of Warren Springs, U.K. and the VDI-Guideline might be a useful base to describe commonly accepted sampling procedures aiming at a standardization of sampling which might be a first step for a harmonization of olfactometric measurements in the different laboratories and countri es. REFERENCES (1) BULLEY, N.R. and D. PHILLIPS (1980). Sensory evaluation of agricul tural odours: A critical review. Can. Agric. Eng. 22, 107 - 112. (2) HENRY, J.G. and R. GEHR (1980). Odour control: An operator's guide. Journal WPCF 52, 2523 - 2537. (3) ROOS, C., J.A. DON and J. SCHAEFER (1984). Characterization of odour-polluted air. In: Proc.Int.Symp., Soc. Beige de Filtr. (eds.), 25-27 April 1984, Louvain-La-Neuve, Belgium, pp. 3 - 22. (4) BAKER, A.R. and R.C. DOERR (1959). Methods of sampling and storage of air containing vapors and gases. Int.J.Air Poll. 2, 142 - 158. (5) SCHUETTE, F.J. (1967). Plastic bags for collection of gas samples. Atmosph.Environm. 1, 515 - 519. (6) SCHODDER, F. (1977T. Messen von Geruchsstoffkonzentrationen, Erfassen von Geruch. Grundl. Landtechnik 27, 73 - 82. (7) CORMACK, D., T.A. DORLING and B.W7J. LYNCH (1974). Comparison of tech niques for organoleptic odour-intensity assessment. Chem.Ind. (Lon don) no. 2, 857 - 861. (8) SCHUETZLE, D., T.J. PRATER and S. RUDDELL (1975). Sampling and anal ysis of emissions from stationary sources. I. Odour and total hydro carbons. APCA Journal 25, 925 - 932. (9) WAUTERS, E., E. WALRAVENS, E. MUYLLE and G. VERDUYN (1983). An evalu ation of a fast sampling procedure for the trace analysis of volatile organic compounds in ambient air. Environm.Monitor.Assessm. 3, 151-160. (10) LACHENMAYER, U. and H. KOHLER (1984). Untersuchungen zur Neuentwick-lung eines Olfaktometers. Staub - Reinhalt. Luft 44, 359 - 362. (11) BERNARD, F. (1984). Simplified methods of odour measurement: Indus trial application and interest for administrative control. Proc. Int. Symp., Soc. Beige de Filtr. (eds.), 25 - 27 April 1984, Louvain-La-Neuve, Belgium, pp. 139 - 150. (12) GILLARD, F. (1984). Measurement of odours by dynamic olfactometry. Application to the steel and carbonization industries. Proc.Int.Symp., Soc. Beige de Filtr. (eds.), 25 - 27 April 1984, Louvain-La-Neuve, Belgium, pp. 53 - 86. (13) MANNEBECK, H. (1975). Tragbare Olfaktometer. VDI-Bericht 226, 103-105. (14) BEDBOROUGH, D.R. (1980). Sensory measurement of odours. In: Odour Control - a concise guide, F.H.H. Valentin and A.A. North (eds.), Warren Springs Laboratories, Stevenage, Hertfordshire, U.K., pp. 17-30. (15) THIELE, V. (1984). Olfaktometrie an einer Emissionsquelle - Ergebnis-se des VDI-Ringvergleichs. Staub - Reinhalt. Luft 44, 342 - 351. (16) DUFFEE, R.A., J.P. WAHL, W. MARRONE and J.S. NADERT1973). Defining and measuring objectionable odors. Internat. Pollution Eng. Congress, Philadelphia, paper no 25a, pp. 192 - 201.
Gui deli nes - The Netherlands: Geurnormering - Odour standard (1983). Ministerie van Volkshuisvesting, Ruimtelijke Ordening en Milieubeheer, directie Lucht, Postbus 450, 2260 MB Leidschendam. - France: Norme experimental - Odour standard, 1st draft (1982). Pol lution atmospheric, methode de mesurage de 1'odeur d'un effluent gazeux, determination du facteur de dilution au seuil de perception - X 43-101. AFNOR, Tour Europe Cedex 7, 92080 Paris. - Germany: VDI-Richtlinie 3881 - VDI-Guideline 3881, part 2, draft (1984). 01factometric method of odour threshold determination. Sampling for odour threshold determination with olfactometers. VDI-Kommission Reinhaltung der Luft. Beuth-Verlag Gmbh, BurggrafenstraBe 4 - 14, 1000 Berlin 30. - United Kingdom: Odour Control - a concise guide (1980). F.H.H. Valentin and A.A. North (eds.), Warren Spring Laboratory, Gunnels Wood Road, Stevenage, Hertfordshire SGI 2BX.
SELECTION AND TREA1MENT CF PANELISTS FOR DETERMINATION OF ODOR THRESHOIDS M. Hangartner Swiss Federal Institute of Technology, Department of Hygiene and Applied Physiology In order to harmonize an odor measurement technique, national guide lines or recaimendations from Germany, France, the Netherlands and the United Kingdom are compared with respect to selection and treat ment of panelists. Different methods of mathematical treatment of threshold data are also presented. 1. INTRODUCTION The task of the working group "odours" in working party one of the COST 681 programme on processing and use of sewage sludge is to give a contribution to harmonize an odor measurement technique. For this purpose the national guidelines or recommendations from Germany, France, the Netherlands and the United Kingdom are compared. The emphasis of this paper lies on selection and treatment of panelists for odor threshold determination. It is well known that sensitivity of men to odorants varies within a large range. By selecting the panelists at one extreme of the sensiti vity distribution, the result can be falsified. However, by chosing a large number of panelists, this effect can be minimized, but this is often not suitable in practise. In the following the different reccmnen-dations are presented. Another source of variation of threshold values is the treatment of panelists, that means comfort, motivation, interaction with panelleader, adaptation etc. These effects can be reduced using a proper detection method. Finally, different methods for threshold data treatment may pro duce different threshold values. For comparison, the following guidelines are reviewed: Germany - VDI Guideline 3881: Olfactometric method of odor threshold determina tion, Fundamentals (Nov., 1983) France - AFNOR Standard: Air pollution - Method of measuring odors from gaseous effluents determination of the dilution factor of the threshold of per ception (1982) Netherlands - Odor standard, Ministerie van Volkshuisevesting (1983) United Kingdom - Odor control - a concise guide prepared on behalf of the Department of the Environment Warren Spring Laboratory (1980)
CONDITIONS OF ODOR THRESHOLD DETERMINATION 2.1 Requirements for the test area Olfactonetric measurement should be undertaken in a roan or area which is kept free frcm odors. There should be an atmosphere of ccmfort and relaxation in the test chamber, which will encourage panel members to concentrate on the testing task and not to be distracted by external sti muli. The test should be carried out at roan temperature and normal humi dity. 2.2 General conditions for test procedure Odor measurements must be carried out with the help of a team leader, who instructs the panelists and operates the measuring equipment. Ccmnu-nication between the team leader and the panel has to be kept to an abso lute minimum. Because of fatigue, the duration of a test series as well as the time of the whole session should be limited. Breaks of at least the same duration as the proceeding test period should be provided. Germany France Nether United lands Kingdcm Panel leader yes yes yes yes duration of 15-30 min 20 min 15 min test series duration of breaks 15-30 min 20 min ? 5 min 30 min 2 test 2 hours time of a test 300 tests/ series of period day 20 tests Table 1: General conditions 3. DETECTION METHODS 3.1 Presentation of odor stimulus 3.1.1 Method of limits The most used method for establishing an absolute threshold in en vironmental studies is the Method of Limits. In its classical form, the stimuli are presented in alternating ascending and descending series, starting at different points to avoid having the subject fall into a rou tine. During this procedure there is a chance that adaptation phenomena may develop. An effort to minimize these effects is for example to use only an ascending series of stimuli. The threshold value for each sepa rate test series is defined as a point in-between the last undetected and the first detected point in the stimulus continuum. A modification of the method of limits is the "up and down" method. A stimulus is presented: if the response is positive, the next lower sti mulus is presented, if it is negative, the next higher is presented and so on. The primary advantage is, that it automatically concentrates near the mean and a considerable number of observations can be saved.
1.2 Method of constant stimuli (Method of frequency) By the Method of Frequency the stimulus range is selected in discrete intervals so that the frequency of positive answers is distributed over the range between 1% and 99%. In general, the frequency of positive res ponses either for an individual or for a group, is cumulatively normally distributed over a geometric intensity continuum. The absolute odor thre shold can then be defined as the effective dose corresponding to an arbi trarily selected frequency of positive responses, ordinarily 50% : ED^: Effective dose at the 50% level. 3.1.3 Signal detection The Signal Detection principle is a determination of the relation ship between hits and false alarms. In determining signal detectability, a stimulus or a few stimuli are presented in random order, alternating with noise. Since sensory impressions resulting frcm the presentation of stimulus versus noise are assumed to be normally distributed over the same intensity continuum and to have the same dispersion, the index of detectability d' for p (hits) minus p (false) indicates the extent to which the two distributions overlap. 3.2 Indication of response 3.2.1 "Yes" or "no" response In the classical evaluation yes-no answers are dependent on the sub jects1 honesty and motivation among other factors. However, yes-no ans wers may be evaluated if they are presented a sufficiently large number of times alternating with blanks. 3.2.2 forced choice technique One method of controlling response perseveration and otter antici pation factors is to use a forced choice response indication based on two or more response categories. In the measurement of odors the panelist has to report the temporal position of positive stimuli in a series of randan blanks. If the concentration is below the threshold, the test sub jects will guess. As the odorant concentration will increase, the rela tive cumulative frequency for identification of the correct sample will be greater. In order to determine the relative odor recognition a cor rection must be made. 3.3 Size of stimulus intervals 3.3.1 Concentration intervals In selecting the stimulus continuum in threshold determination, the relation between just noticeable difference in relation to the intensity of stimuli is of interest. In accordance with Weber's law this quotient is assumed to be a constant. Therefore it would appear best to determine absolute thresholds on an intensity continuum in the form of a gecxnetric progression. 3.2.2 Time intervals Because of adaptation processes the exposure time until reaching a decision should be limited. Also the interval between two stimuli must be observed.
f......... ■ indication of yes/no forced ch. forced ch. yes/no response dilution steps 1.5-23 * 21.6 exposure time <15 sec 10 sec <15 sec 20 sec stimulus interval 15 sec 3 min 1 min 1 min Table 2:Detection method, indication of response and size of stimulus intervals. 4.1 Definition of odor threshold By convention the individual odor threshold is that concentration which is just perceived by the subject in 50% of the cases in which it is presented to him. The group threshold is the concentration that is just perceived by 50% of the panel members. 4.2 Evaluation using the geometric mean The point of change is determined for every series of dilution eva luated. It is defined as the geometric mean of the dilution of the last negative and the first positive answer. The arithmetical mean and its standard deviation are calculated frcm the logarithms of the points of change. 4.3 Graphical evaluation The characteristic curve of the odor threshold is used. The rela tive cumulative frequency of positive answers is calculated for each odorant concentration and graphically plotted, while for odor concen tration a logarithmic scale is used. The odor threshold can be obtained frcm the resulting curve as the 50-percentile and so can the associated 16- and 84-percentiles. 4.4 Probit analysis If the odor sensitivity is normally distributed over the logarithm of the odor concentration, the characteristic curve of the odor threshold is a gaussian curve. This curve is converted into a straight line using the probit transformation. The analyses can be carried out graphically on probability paper or by transformation of the relative cumulative fre quency by using a table function and calculating the regression lines. The odor threshold and the 16- and 84-percentiles can be determined frcm the results.
Germany France Nether United lands Kingdan detection method limit limit limit limit randan/ up & down ascend. random ascend.
Germany France Nether United lands Kingdom mathematical gean.mean gecm.mean gecm. gecm.mean procedure graphic. graphic. graphic. graphic. probit-analysis Table 3; Mathematical treatment of threshold data 5. SELECTION OF PMALISTS 5.1 Requirement for panelists Panlists are required to have the following qualities: - sensitivity: subjects with specific anosmia or hyposmia must be ex cluded. - physical condition: subjects whose sense of smell is temporarily im paired by desease must be excluded. - reliability: subjects must be able to reproduce accurate results con sistently. - honesty: subjects must exactly say what they perceive. Germany France Nether United lands Kingdom Sensitivity "normal" 5 reference not too actual odor sense, age odors range good, not or key carp. 18-50 y. 1:1000 too bad Physical condition quest. quest. Reliability repeated measures with H2S Honesty 20% errors Table 4: Selection of panelists Olfactory sensitivity for one individual varies about factor three due to climatological, physiological, environmental reaons etc. The sensory sensitivity also varies from odorant to odorant. So it is difficult to select a panel with a sensitivity distribution similar to that of the population. The preferred method in the United Kingdom for screening panelists uses the actual odor to be tested as a key component. In France selection is carried out on the basis of the threshold for five standard odor ants. In Germany a "normal11 sense of smell is requested of persons between the age of 18 and 50 years, in the Netherlands no exact specifi cations are given. Anyway, an extreme clustering around the mean or to wards the extremes has to be avoided.
The physical condition is checked by the panel leader using a question naire or simply by asking the test persons according to the guidelines of Germany and the United Kingdom. Only the German guideline tests reli ability of panel members by repeated measurements with the same odorant. Problems of honesty are minimized by forced choice technique (France, Netherlands). In the German guideline persons with more than 20% of errors in more than three test series are excluded. 5.2 Panel size The extent to which a panel constitutes a representative sample of the population depends directly on the numbers of panel members. For practical reasons a cxxrpromise must be sought between costs and the representative ness of the result, and this depends on the question to be answered: basic measurement e.g. emission standards or only comparative measurements, e.g. odor abatement efficiency. Nether Germany United France lands Kingdom basic measure 8-15 10 6-86-8 ments comparative measurements > 4 Table 5: Panel size. 6. CCECLUSICNS - There is more or less agreement in all guidelines about general back ground conditions. - The limit method is proposed as detection method in all guidelines. The indication of response is either yes/no or correct/incorrect. The latter, forced choice technique, may certainly give lower odor thres holds. - The mathematical treatment of data will produce only slight differences in the threshold values. - For the panel size different members are given. 8 people appears the right size for the panel. - Selection of panelists is the most difficult question and large varia tions of threshold data are expected due to this problem. No generally accepted procedure exists and only vague recanmendations are given in the guidelines. A possible solution will be the evaluation of the sensi tivity distribution of a large panel (>25) of the actual odor to be tested, and screening the panel members according to their position in the distribution. However, this procedure might not be suitable in practice.
VDI GUIDELINES ON ODOUR PROBLEMS Dr. V. Thiele Landesanstalt fur Iiranissionsschutz des Landes NRW D-4300 Essen, FR Germany Three working groups of VDI are mainly engaged in pro blems of odour determination and odour assessment. The working groups are titled: -"Odorous Substances" -"Application of Olfactometric Methods and Performance Characteristics" -"Dispersion and Odour Concentration in Ambient Air". The titles characterize the subjects which are treated in these working groups. At present the groups establish guide lines to solve odour problems. The guidelines are in diffe rent states of development. 3881-1: Olfactometry - Odour Threshold Determination -Fundamentals -2: Olfactometry - Odour Threshold Determination -Sampling -3: Olfactometry - Odour Threshold Determination -Olfactometers Types 1158 and TO-4 -4: Olfactometry - Odour Threshold Determination -Instruction for Application and Performance Characteristics 3882 : Assessment and Effects of Odours - Intensity and Hedonic Tone 3883 : Assessment and Effects of Odours - Measurement of Annoyance by Means of Interviews 3940 : Odour Determination in Ambient Air by Inspection Panels 3781 : Odour Dispersion and Odour Concentration in Ambient Air In the following some aspects of the guidelines are given in detail. Guideline VDI 3881 consists of four parts. The drafts of parts 1, 2, and 3 were published in the VDI handbook. Part 4 is in preparation. The draft of part 1 was already revised. The new version will be published in a few months. The most important result of the revision is the definition of odour concentration expressed as odour units per cubic meter
(GE/m**3). According to this definition one odour unit is the amount of odorants in one cubic meter of air at odour thres hold level. The new definition is a real concentration and gives a better form of input parameter for dispersion models. On the basis of guideline VDI 3881 parts 1, 2, and 3 ringtests were carried out with different odorants. The results can be summerized as follows: -The dispersion of results varies and depends on the compo sition of the participants and on problems of sampling and preparation of odorous sample. Lower dispersion is obtained when results with obvious errors in application of guide lines or with large deviations from mean value are excluded. -Participants of the Netherlands get systematically lower threshold values than the others. The reason has to be investigated. -All findings of the ringtests lead to the conclusion that it is possible to determine odour thresholds which do not differ by more than factor 10. At present another ringtest is in preparation. This test will be carried out in summer 1985. The French collegues will also participate in this test. Experience of all ringtests will be reported in part 4 of guideline VDI 3881. Guideline VDI 3882 deals with the determination of odour intensity and hedonic tone. The members of the working group "odorous substances" assume that odour threshold and odour concentration are insufficient for the characterization of odorous perception. They recommend to judge the odour inten sity and the hedonic tone by category estimation. Moreover, it is their opinion that the odour determination with olfacto meters is not suitable to assess odour in ambient air. There fore they are preparing two guidelines dealing with these problems. Guideline VDI 3883 gives instructions on the regis tration of nuisance by interviews with nearby residents of emitting plants or inhabitants of industrial areas. Addition ally guideline VDI 3940 describes the determination of odour in ambient air by inspection panels based on the following idea: During constant conditions as to the class of weather, wind speed, and wind direction each local point is charac terized by a frequency of odour perception representing the probability to perceive an odour. The situation at a local point will be have to determine the portion of a year with a frequency of odorous perception greater than 5 % in a random test. Both guidelines, VDI 3882 and 3940, should give corres ponding results. Guideline VDI 3781 part 5 completes the complex of odour determination and judgement with the calculation of disper sion models. The calculation methode and odour determination by panelists should give comparable results. The following summery can be given. Odour measurements with olfactometers is only a small part of the whole field of odour determination in ambient air and the measurement of odour nuisance must be approached in the near future with appropriate urgency.
the panel leader on a panel of light in a separate signal box. The panel leader records the judgements and calculates by Beans of a statistical procedure, the averaged panel value termed E D . This term denotes Effective £osage at the 50 percent level, i.e. thlt dilution level at which 50 percent of the panel would and 50 percent would not detect odour of the diluted sample. The dilution is denoted by the dilution factor. For instance: ED = 1000 means that one litre of the odorous air must be diluted with 1000 litres of non-dodrous air to reach the panel threshold termed ED . With the olfactometer in use it is possible to measure dilutions between approximately 10x up to approximately 30.000x. 3. APPLICATIONS In general the main sources of odour emission in Norway are fish meal plants, pulp and paper mills, and plants for the treatment of sewage sludge and waste water. Investigations have been carried out in these and other branches of the industry, i.e. the food industry. Each case may provide features which influence the olfactometric measurements, often demanding special sampling techniques and inter pretations. In the following some of the problems and experiences will be pointed out by means of examples from sewage treatment and fish meal plants, showing the use of olfactometry for obtaining satisfactory odour reducing results. A couple of years ago an investigation was carried out to evaluate the efficiency of odour reducing processes based on different principles, such as chemical scrubbers, soil bed filters, activated carbon filters, iron oxide filters, and combustion. Samples for the olfactometric measurements were taken in different positions in the installations and the odour reducing efficiency was calculated as the ratio between the recorded ED value at the outlet and inlet of the purification steps: nfilter:
information provided by the description of the odour is indispensable in the evaluation of the purification effect. When operated incorrectly the odour of the sample will be described as "chemicals, faint chlorine" or worse, “typical chlorine". Quite often such informations are used as a basis for adjustments and optimizing in chemical scrubber management. Measure of odour reducing efficiency in iron oxide filters is another esample of how olfactometry may contribute to the optimization of an odour reducing method. In some Norwegian sewage sludge and wastewater treatment plants iron oxide filters have been installed with success. The filter consists of mixed wood chips and iron oxide, and the odorous compounds are oxidized in the filter. The total odour strength was measured in such a filter where the air frpm sludge tanks providing an offensive odour was finely dispersed at the 3 m bottom of the filter box (total filter volume: 3 m , containing 300 kg iron oxide). It was observed that the purifying efficiency was strictly dependent on the gas flow in the filter, and that relatively high flow rates gave the highest efficiencies. This is indicated in the curve below. n filter (\) 90 •-80 --70 •• 60 -50 - — I--------1--------1--------1 esKfilaBBgRt, o£ a "Bsy" reducing istkQfl :
50 100 150 200 250 (m3/h) Gas flow Figure 1. Odour reduction efficiency in iron oxide filter expressed as a function of the gas flow through the filter. This is an example of how present processes can be combined to obtain a better odour abatement system, and how olfactometric measurements are useful by evaluation of the efforts made to improve the odour reducing
efficiency. By measurements of total odour strength in a treatment plant the ED values pointed out the sludge press and dewatering process as the predominant odour sources of the plant. In the venting air from this position extremely high ED values were recorded. This air was led through a carbon filter for odour reduction. Olfactometric measurements at the filter revealed poor odour reducing efficiency. It was observed that odour compounds were not destroyed in the filter. They only restrained until the carbon became saturated, and thereafter evaporated into the outlet air contributing to the odour strength. The filter capacity was obviously too small for the heavy load. Attempts to reduce the odour strength before the filter did not succeed, until the air was led through a container filled with saturated lime slurry (pH = 12-14). The slurry was part of a precipitation process in the plant. Dispersion in the alkaline slurry extensively reduced the odour strength of the air, resulting in sufficient capacity of the carbon filter also when handling heavy loads of sewage sludge. Since then the carbon filter has worked well, within the limitation of such filters in general. Neither is it observed signs indicating reduced precipitation properties of the lime slurry. Measurements of total odour strength in combustion processes imply sampling challenges. Beside the chemical scrubber process, combustion of odorous air is the best odour reducing method. The disadvantage of this process is the high energy costs. Treatment at apropriate conditions, however, will destroy the odorous compounds extensively. Temperatures about 850 C and contact time up to 3 seconds are reported (2,3). Olfactometric measurements in combustion processes involve certain sampling problems caused by the temperature difference between inlet and outlet. The humidity of outlet air must also be taken into consideration. Problems may occur when hot outlet air is sampled at low temperatures. In most such cases sampling is impossible without special arrangements. Such conditions are present during odour measurements in fish meal plants with combustion as the odour reducing method. The largest problem turned out to be the temperature differences between outlet air (85-220 C) and outdoor temperatures (0-15 C), causing condensation. The dew point of the outlet air was calculated, and experiments were carried out with dilution of the outlet air to prevent condensation in the sampling bags. Condensation was prevented by diluting the outlet air 5-150 times with dry, purified N gas. Comparison of N -diluted and undiluted samples revealed large differences in ED value. In samples demanding a high degree of dilution to prevent condensation, the measured odour strength was up to 5 times higher than in the undiluted corresponding samples. Samples demanding less dilution showed less deviating results. 4. CONCLUSIONS In the attempt to minimize odour emission, olfactometric measurements of total odour strength give useful informations about the odour reducing efficiency of different processes as a function of parameters like dosage of chemicals in scrubbers, humidity and temperature in packed filters, flow rates, etc. Olfactometric measurements also point out the main odour sources of the plant. From a set of olfactometric data combined with other essential
parameters, efforts can be made to improve odour reducing processes within their limitations. Better odour reducing efficiency can be obtained by appropriate management of the process, and energy costs may be reduced by discriminative venting of process operations contributing with high concentrations of odorous compounds. However, attention must be paid to the sampling procedures in processes involving high temperatures and high degrees of humidity. Measurements indicate that such conditions may influence on ED values to some extent. REFFERENCES (1) DRAVNIEKS, A. and PROKOP, W.H. (1973). Source emission odour measurement by a dynamic forced choice triangle olfactometer. Air Poll. Control Assoc. Paper, 73-276. (2) PETTIT, C.G. (1959). 20 years of sewage sludge burning at Barberton, Ohio. J. San. Eng. Div. Amer. Soc. Civil Engr. 85SA6, 17. (3) LABOON, J.F. (1961). Construction and operation of the Pittsburgh project. J. Water. Poll. Control Fed. 33, 758.
DISPERSION MODELS FOR EMISSIONS FROM AGRICULTURAL SOURCES G.-J. MEJER and K.-H. KRAUSE Institut fiir landtechnische Grundlagenforschung der Bundesforschungsanstalt fiir Landwirtschaft Summary The aim of dispersion models is the prediction of atmospheric dilution of pollutants in order to prevent or avoid nuisance. Established dispersion models, designed for the large scale of industrial air pol lution have to be modified to the small scale of agricultural pol lutions. An experimental setup is described to measure atmospheric dilution of tracer gas under agricultural conditions. The experimental results deliver the data base to identify the parameters of the models. For undisturbed airflow modified Gaussian models are applicable. For the consideration of obstacles more sophisticated models are necessary. 1. INTRODUCTION The aim of dispersion models is to develop reliable methods for calcu lating the atmospheric dilution of airborne pollutants under practical conditions. One application in agriculture is the determination of that distance, at which i.g. odouriferous pollutants of an animal farm are diluted in the atmosphere to a concentration below a certain threshold, in order to allow the farmer a profitable production and likewise to prevent odour nuisance from the neighbourhood. Another application is the prediction of the effectiveness of changes in the emission source configuration, in order to reduce the odour nuisance in the existent vicinity. That could help to avoid expensive misinvestments. In air pollution control it is useful! to subdivide this large problem into three main divisions /1/, fig. 1:
the emission; this is the entrance of the airborne pollutants into the open atmosphere. The local position of this entrance is the emission source, - the transmission, including all phenomena of transport, dispersion and dilution in the open atmosphere, - the immission; this is the entrance of the pollutant into an acceptor. As we are regarding odoriferous pollutants, the immisson is their entrance into a human nose. About air pollution from industrial emission sources, i.g. S02 from power plants, a wide knowledge is available, including sophisticated methods of emission measurement, atmospheric diffusion calculation and measurement of immission concentration in the ambient air. In most countries we have complete national legal regulations, concerning limitation of air contaminent emissions, calculation of stack height and at least evaluation and determination of maximum inmission values. Within this situation the question arises, whether these wellproved methods and devices are suitable for agricultural odour emissions from agricultural sources too. It is well known that all calculations and values, established in air pollution control, are based on large sets of data, obtained by a multitude of experiments and observations. The attempt to apply these established dispersion models to agricultural emission sources, leads to unreasonable results. A comparison in table 1 shows that the large scale values of industrial air pollutions, on which the established dispersion models are based, are too different from those in agriculture. In order to modify the existing dispersion models or to design other types of models, we need the corresponding sets of observations and of experimental data, adequate to the typical agricultural conditions. There are already a lot of investigations to measure odour at the source and in the ambient air. But we all know about the reliability of those measurements and about the difficulties to quantify these results adequate to a computer model calculating the relation between emission and immision depending on various influences and parameters. So we decided to supplement the odour measurements by tracer gas measurements easy to realise with high accuracy. The aim is to get the necessary sets of experimental data for the modification of existing dispersion models for agricultural conditions. 2. INSTRUMENTAL 2.1 EMISSION the published guideline VDI 3881 /2-4/ describes, how to measure odour emissions for application in dispersion models. Results obtained by this method have to be completed with physical data like flow rates etc. As olfactometric odour threshold determination is rather expensive, it is supplemented with tracer gas emissions, easy to quantify. In the mobile tracer gas emission source, fig. 2, up to 50 kg propane per hour are diluted with up to 1 000 m3 air per hour. This blend is blown into the open atmosphere. The dilution device, including the fan, can be seperated from the trailer and mounted at any place, e.g. on top of a roof to simulate the exaust of a pig house or in the middle of a field to simulate undisturbed air flow. 2.2 TRANSMISSION For safety reasons, propane concentration at the source is always below the lower ignition concentration of 2,1 %. As the specific gravity of this emitted propane-air-blend is very close to that of pure air (difference less than 0,2%) and as flow parameters can be chosen in a wide range, we assume
that the atmospheric diffusion of this blend is equal to that of odour-polluted air from agricultural emission sources. In a first step the dispersion in an undisturbed atmospheric flow was measured. In a second step the dispersion in an atmospheric flow with obstacles is designated. In a third step the dispersion in the neighbourhood of real farm buildings is intended. 2.3 IMMISSION 2.3.1 SPATIAL DISTRIBUTION OF TRACER GAS CONCENTRATION To obtain multiple sets of experimental dispersion data, in each experiment 50 samples of tracer-polluted ambient air downwind in the plume of the propane emission source are taken by 10 sample units, distributed in the field, see fig. 3. Each unit carries five glass cylinders, filled with +9 +6 V /
Fig. 3. Experimental setup for field measurements without obstacles. water and connected with small tubes to the different immission sample heights, see fig. 4. Passing a throttle and a solenoid valve, the water flows down by tne force of gravity into a tank, thus sucking the sample
emissions. Fig. 4. Sample unit, a sample in take e water pipe b sample conduct f tank c valves g throttle d glass cylinders h solenoid valve.
Geruchsschwelle = odour threshold Mittelwert = mean value Ref.: MEDROW and JORGENS /6/. As grade and frequency of fluctuation are also dependent on this distance, the effect of odour concentration fluctuation must be quantified in this range by reliable measurements and taken into consideration in the
turbulent diffusion; at the ground total reflection is assumed. For the case that the concentration at any point in space is independent of t and that the diffusivities are independent of x, y and z the simplified diffu sion equation of the K-therory /8/ becomes 9C - 3"C4. V 32C
U 3x ' Ky W Kz Tz? (3.2). With the boundary conditions at z = 0: = 0 (3.3a), -*°°:C+0 (3.3b), the condition of continuity at x > 0: f f U C dy dz = C0 V (3.3c), 0 -00 the product C0 ft of the source concentration C0 and the volume emission rate V stands for the strength of the point source, and with the initial condition
AK FfGgmI1,294 0,718 0,241 0,662 0,419 II 0,801 0,754 0,264 0,774 0,369 III-1 0,640 0,784 0,215 0,885 0,282 III-2 0,659 0,807 0,165 0,996 0,223 IV 0,876 0,823 0,127 1,108 0,205 V 1,503 0,833 0,151 1,219 0,089 Ak F F0 f G Go 9m50,25 194 0,72 0,69 48 0,59 0,40 4 1,19 354 0,67 0,60 42 0,74 0,28 3 0,82 198 0,74 0,09 12 1,11 0,22 2 0,94 168 0,76 0,08 12 1,28 0,11 Table 2. Coefficients of diffusivity dependent on stability classes after Klug and Turner; m stands for the exponent in the power law of wind velocity. However, we must keep in mind the limitations of this approach, especially the transfer of consistent sets of dispersion parameters to the propagation of air pollution in the vicinity of a source. The Gaussian plume formula should be used only for those downwind distances for which the empirical diffusion coefficients have been determined by standard diffusion experiments. Because we are interested in emissions near ground level and immissions nearby the source, we use those diffusion parameters which are based on the classification of Klug /12/ and Turner /13/. The parameters are expressible as power functions, Oy(x) = F xf and az(x) = G x9 after Klug (3.6a,b), tfy(x) = (F + Fx)f and az(x) = (GQ + Gx)9 after Turner (3.7a,b). The parameter classification after Klug is determined by six stability classes (with the German abbreviation AK for Ausbreitungsklasse), reaching from extreme stable (AK I) to extreme labile TAK V). In tRe Turner stability scheme AK 5 denotes extreme stable, AK 2 extreme labile, see table 2. An estimate of the stability can be made from synoptical observa tions of solar radiation, cloud cover and wind velocity /14/. With the parameters after Klug equation (3.4) becomes C(x,y,z) = ax"(f+9^exp(-bx"2f) [exp(-d0x"2g)+exp(-d1x"2g)] (3.8), wherein - - C0V k ya w (z-H)2 ^ (z+H) a ' TrTOFE • b ■ ■JT • do = -Z IP '- • d1 = ~75*~
In the mean direction, y = 0, the first exponential expression gives (exp(-x"2f))b = 1 for b = 0 and all x > 0. Equation (3.8) is reduced to C(x,y,z) = ax‘(f+g) [exp(-dQx"29) + expt-d^'29)] (3.10). The essential behaviour of the transport model in the vicinity of the source can be discussed by the equation (3.10), searching for the maximum concentration in direction of the centerline, dC/dx = 0. From equation (3.11) the distance x > 0 is computable at which maximum concentration occurs: 2gd d, d-d, 9d-d, 3.11). At ground level, z = 0 and dQ = d-j = , xmax becomes
by iteration methods. If it is assumed that the exponential expression is < 1, equation (3.14) reduces to that means
demonstration, see fig. 7. Gas is released at the emission point x = 30 m, y = 15 m and z = 2 m at a volume rate of 400 m3/h; this volume rate is required in summertime per unit of livestock (in German: GroBvieheinheit, 1 GV = 500 kg live weight) for ventilation of pighouses; Tt is an average measure. Fig. 7 shows the isopleths of the dimensionsless concentration ratio C/C at the immission level z = 3 m. The mean wind speed U is 2,6m/s. The arrows give the range of the altering wind direction; the mean wind direction is 161 degree. At the source we have C/C0 = 10\ A maximum of nondimensional concentration is found in the downwind distance of about 30 m from the point source. 60 ^ 30 15 0 15 30 £5 m 60
Fig. 8. Isopleths of concentration ratio C/C0 in the y-z-plane at x = 15 m (at the arid points measured data)
Fig. 10. Plot of downwind concen tration ratio C/C0 against distance; C/C0 is calcu lated by means of the
In the special case shown here, the Gaussian plume model does not predict the location of the maximum concentration in agreement with the experiment, but it is appropriate to determine the concentration decay in downwind direction. That what happens between the point source location and the maximum location is of accademic interest only. A question for practical purpose is how we can get information about the maximum location, where from the model is realistic. From equation (3.13) we can deduct a rough approximation of the location where maximum ground-level concentration occurs. It is argued that the turbulent diffusion acts more and more on the emitted substances, when the distance from the point source increases: therefore the downwind distance dependency of the diffusion coefficients is done afterwards. If we drop this dependency, equation (3.13) leads to Xmax = 34,4 m for AK = I (curve a) and xmax = 87,7 m for AK = V (curve b), what is demonstrated in fig. 11. The interpolated ranges of measured values are lined in. Curve a overestimates the nondimensional concentration maximum, but its location seems to be correct. In the case of curve b the situation is inverted. C urve c is calculated with the data of AK = II. The decay of the nondimen sional concentration is predicted well behind the maximum. Curve d is produced with F - 12,1, f = 0,069, G = 0,04 and g = 1,088. The ascent of concentration is acceptable, but that is all, because there is no explana tion of plausibility how to alter the diffusivity parameters. Therefore it must be our aim to find a suitable correction in connection with the meteorological input data. o 0
Fig. 11. Calculated C/C0-distributions for different diffusivity parameters, curve a) F = 1,294, G = 0,241, f = 0,500, g = 0,500
velocity field; another one is given by the fluctuation u' in form of the turbulence degree Tu = /iT^/U. For y = 0 and z = 0 the concentration relationship is modified, J H2a* = iToT (1 ' exp x>0 °x H2 °z
x y z z Herein we have assumed < 1 (4.2) ax may be related to the turbulence degree Tu; in the used example Tu = 0,243. By a* the concentration of equation (4.1) is reduced. We will see if there exists any usable form ax = f(Tu,x). 5. CONCLUSIONS
(9) HEINES, T.S., L.K. PETERS: An analytical investigation of effect of a first order chemical reaction on the dispersion of pollutants in the atmosphere. Atmospheric Environment 7 (1972) S. 153-162. (10) CARSLAW, H.S., J.C. JAEGER: Conduction of heat in solids. Oxford: Claredon Press 1959. (11) SCHONBUCHER, A., V. SCHELLER: Ausbreitung von Abgasfahnen. Chem.-Ing. Techn. 53 (1981) Nr. 5, S. 320-334. (12) KLUG, W.: Ein Verfahren zur Bestimnung der Ausbreitungsbedingungen aus synoptischen Beobachtungen. Staub-Reinhalt. Luft 29 (1966) Nr. 4, S. 143-147. (13) TURNER, D.B.: Workbook of atmospheric dispersion estimates. Public Health Service Publication No. 999-AP-26. Washington: Department of Health, Education and Welfare 1969. (14) RdErl. d. Ministers fiir Arbeit, Gesundheit und Soziales vom 14.4.1975, Verwaltungsvorschriften zum Genehmigunqsverfahren nach §§ 6, 15 - Bundes-Immissionsschutzgesetz (BImSchG) fiir Mineraloraffi-nerien und petrochemische Anlagen zur Kohlenwasserstoffherstellung (Raffinerie-Richtlinie) Ministerialblatt fiir das Land Nordrhein-Westfalen, Jahrgang 1975, Nr. 65, S. 996-1007.
Fig. 1. Air flow equipment in the Prosser olfactometer Valve VI controls the sample flow rate Valves V2 and V3 are in the same housing and switch flews between the low (L) and high (H) ranges. In the high range, fan 2 provides a second-stage dilution of 100X, which is controlled by valve V4 Fig. 2. Laminar flew meter connected to sample inlet port of Prosser olfactometer
Fig. 3 Variation in the sample flew rate with the angle of rotation of the sample control valve. The integers within the graph refer to the number of rotations. Hie y axis is shown on a log scale. Hie parallel, dashed lines, show the eleven flow rates corresponding to the eleven dilutions
imLQPHEMT&.m.THE ASSESSMENT OF ODOURS FROM SLUDGES S. J. Toogood Water Research Centre, Stevenage and J. Diaper, School of Environmental Science, University of Bradford SUMMARY 1. Odours Iron Norbfoint.Aynges
The remarkable sensitivity of the human nose to detect a large number of chemicals by smell means that traditional odour monitoring methods are costly either in capital (chemical analysis) or labour (olfactometry). Not least among the problems in achieving worthwhile results is the difficulty in representative sampling. This difficulty is The major difference however is that industrial processes are normally carried out in enclosed vessels, whereas open tanks are the rule at a sewage works for storage and for much of the processing This means that instead of an emission of odour from a chimney whose flow rate can be measured and quality analysed, the odour diffuses from the surface of
In any case, covering may be impracticable for other reasons. Many processes as they are currently designed depend upon at least visual access by operators for process control, and in other instances the production of odorous chemicals such as hydrogen sulphide can be accompanied by the formation of methane, giving a potential fire or explosion hazard. 1.1.1. Odours .from the Spreading of Sludge and Slurries on Land The chimney, originally devised to increase draught through fires and to provide smoke extraction, has found extensive use in many industries to aid the dispersion of odour. The extra height gained by the point of emission is frequently enough to give the extra dilution required to reduce the risk of odour nuisance at even relatively nearby properties. This is an option that is not open to the farmer or the sewage works operator. The cost and practicability of enclosing the processes used in sewage treatment varies considerably, but in the disposal of sludge or animal slurries to land there is never a realistic option that the dispersion of odours once transferred to the gas phase might be effectively controlled. For the prevention of nuisance therefore there are two possibili ties. First, the formation or release of odorous chemical species can be discouraged. In practice this usually means the prevention of reducing conditions (negative redox potential) and possibly the prior removal of certain key compounds. Second, the time of contact between the sludge/ slurry and the air can be reduced, for example by ploughing in or sub-surface injection, and the act of spreading can be timed to coincide with favourable atmospheric conditions. These two approaches can of course be used in combination. Both approaches naturally add to the cost of sludge disposal, and for the sewage works manager add to the risk that farmers might be less willing to accept sludge to land, causing a greater problem still. For the fanner, sewage sludge can be a useful source of cheap nitrogen, though of unspecified strength, and also of much needed soil structure, but the imposition of no-grazing periods after application can add to the cost taken as a whole. A further problem, especially for farmers with arable crops is that the demand for soil nutrients and the practicability of spreading and ploughing in are seasonal, whereas a sludge and slurry are produced at a more or less constant rate. In the case of slurries, seme form of storage is inevitable, and commonly takes place in open pits. Scxne digestion and therefore stabilisation takes place during storage, reducing the capacity of the slurry to cause odour nuisance, and as long as the surface crust is not disturbed, little odour results. It is the emptying of slurry pits that gives rise to the release of odour. 2. P rin ciple, Sources, of Odour at .^ weg e .lreatment Works
The capacity of sewage and sewage sludge to produce odour varies markedly during passage through a treatment works, depending on a nunber of factors, including redox potential, pH and temperature. Nevertheless generally it is the operations that give rise to increased contact between the liquid phase and the surrounding air that give rise to
problems. Odour problems are commonly associated with: ♦Septic sewage arriving at works •Industrial discharge arriving at works •Inlet works •Primary treatment •Storm tanks •High rate filters •Conventional filters •Activated sludge process •Transfer of sludge •Treatment of sludge •Storage of sludge Four of these operations are more important than the rest: septic sewage, high-rate filters, sludge transfer and sludge treatment, especially filter pressing, account for the majority of instances of serious nuisance. Of the three processes listed involving sludge two are broadly analogous to land disposal: sludge transfer carries the risks associated with spreading, while the sludge storage corresponds more to sludge on land after spreading. The general conclusion from considering all these factors is that when financial resources for remedial measures are at a premiun, it is particularly important to be able to judge the potential benefit in particular cases, as well as to assess the risk from various operations and to ensure that designs of new works or of operating procedures take properly into account these risks. 2. PROBLEMS IN APPLYING CONVENTIONAL ASSESSMENT TECHNIQUES TO A concept which has proved useful in assessing odours from contained sources is that of emission rate, defined: E = FxD where: E - Emission Rate - Related to complaints and is the input to dispersion modelling and to simpler empirical formulae D - Dilution Number - The number of times the odour must be diluted before 50% of people cannot detect it F - Volume Flow Rate of the odorous air When dealing with non-point sources this formula is difficult to apply, because although values of D can be obtained at specific points around the source, it is extremely difficult to measure or estimate values of F which correspond to them. Measurements around a sludge storage tank in Yorkshire Water Authority, shown in Table I. illustrates the difficulty of prediction. Ideally an ’absolute’ method, capable of measuring F accurately should be used; since this is not possible, an alternative approach is to quantify in a reproducible way the potential of the liquid phase for
Table I. Strength (Dilutions to Threshold) of odour samples from Knostrop sludge tanks Wind velocity (m/s) Sample point 0.41.83.6 20 mm above sludge level 80 195 80 1 m above sludge level 80 128 63 4 m above sludge level 63 80 40 6 m above sludge level 80 100 46 and 20 m from tank producing odours, and to calibrate this to real applications by experience. A method which estimates the odour potential from a sludge or slurry could be used to develop empirical relationships with complaints received during e.g. spreading. Additionally such a method can be used to evaluate treatment processes. A further advantage of this approach is that it deals well with the logistics of sampling and measurement. Taking voluminous odour samples from the vicinity of a liquid surface over an open tank requires experience and time if a representative sample is to be obtained, since the act of sampling disturbs the air being sampled unless the sample flow rate is infinitely slow. Not only this, but it is not unusual for the site being assessed to be a considerable distance from the place where the olfactometry or chemical analysis is to be carried out. Storage of samples in transit results in a somewnat unpredictable change in the odour measurement. Results reported by Gillard^^ show attenuations between 40 and 55% for sane samples over a 24 h storage period, and more remarkably, an augmentation of 139$ for one sample after 8 days storage. Though sewage sludge does change with storage, the effect is minor if the sample is refrigerated. The odour sample can then be extracted at the place of analysis, and used directly. 3 . m .MEASUREMENT QF 'QPfflft POTENTIAL’
In order to make the approach practicable it is necessary to introduce the concept of ’odour potential1. By this we mean the propensity of a liquid or slurry to release odorous substances to an atmosphere; in other words, a measure of the amount of potential odour available for future release. Qg.a Sludge To enable odour potential of different sludges to be compared in the laboratory, a method was required which could be carried out under standard conditions. An initial method was to pass nitrogen (or air)
over sludge spread in the bottcm of a Buchner flask and to collect the resultant odorous gas stream in a 40 litre odour sampling bag made of ’Tedlar1 for use in olfactometry. However, resulting odour strengths were lew because large volunes of nitrogen (or air) were required. A further difficulty was the drying of the sludge at its surface, thus changing the sludge character. Attempts to overcome this by passing humidified air resulted in water accunulation and sometimes flooding of the sludge surface. alternative method was required to increase the concentration of odorous molecules in the vapour phase and so produce a sample which could be used directly for olfactometry. The stream of air was therefore bubbled through, rather than blown over, the sludge. Transpiration methods of this kind have been used to determine vapour pressure and activity of volatile in gas bubbled
The three samples which follow illustrate work done at the University of Bradford on the assessment of odours from sludge treatment disposal^). 4.1. Odour Potential of Ex-works Sewage Sludge Odour measurements were made on routine ex-works sludges collected
Results for (ii) and (iii) are shown in Table III, which records the range of odour potential for samples collected over a one year period. Mean values for odour potential are also shown, these having been calculated from approximately 15 measurements. From Table II the seasonal variation in odour problems is clearly apparent. This effect is most marked with hunus sludge as the proportion of viable bacteria in the sludge increases markedly during the spring, and coupled with the rising sludge temperature greatly increases the kinetics of the biochemical reduction of sulphur and nitrogen compounds. Table II. Variations with sample date of odour potential of humus sludge Odour strength (D) (Dilutions to Threshold) Sample Range Mean Monthly dates mean 22.1 5k-31.6k 15.8k ) 5.21.3k-20k 15.8k ) 15.8k 12.2 10k-20k 15.8k ) 19.2 12.6k-20k 15.8k ) 26.2 31.6k-40k 31.6k ) 25k 12.3 5k-20k 15.8k ) 26.3 20k-80k 50k ) 2.4 80k-200k 126k ) 126k 9.4 200k->250k 200k ) 14.5 250k->250k >250k 21.5 250k->250k >250k j >250k Table III. Inter-works variation in sludge odour potential Odour strength (D) (Dilutions to threshold) Fresh sludge 3-week old sludge Sludge and source (up to 24 hours old) Range Mean Range Mean Esholt 6.3k - 8k 8k 6.3k 8k primary Esholt 4k-250k 63k 630-6.3k 2k humus Knostrop 5k-15.8k 10k 6.3k-12.6k 10k primary/htmus mix Owlwood 320-5k 2.5k 12.6k-63k 40k surplus activated Owlwood not available 32k-80k 63k primary/surplus activated mix
From Table III, the offensive nature of untreated sludge can again be seen. It is also clear that the effect of storage is most marked with secondary sludges. The results in these tables are illustrative of the kind of assesanent of sludges that can be made, and the figures are in a form that can be used directly as an indication of the relative nuisance likely to arise during disposal. The effect of stabilisation is shown later. 4.2. Scheduling Sludge Disposal - The Effect of Storage Time When sludge samples were stored over a 32 day period the results showed a characteristic double peak in odour potential. Examples are shown in Fig. 2 for two mixed and one hunus sludge from sewage treatment works in the Yorkshire Water region. An initial increase can be seen to a maximun at 2-4 days, followed by a decrease before a second maximum at 8-10 days. The pattern indicates that the age at which a sludge is removed for disposal can be critical if odour nuisance is to be minimised. The tests indicate for example that Esholt hunus sludge should be stored for at least four days before disposal, and that longer storage of, say, three weeks is appropriate for the Knostrop sludge. Disposal while fresh may conversely be the best strategy for the Owlwood sludge. Field observations do indeed indicate that Owlwood sludge consistently gave the most odorous conditions during spreading. 4.3. The Effect, of Lime stabilisation.,on.Odours
The stabilisation of sludge by the addition of lime is favoured in some places. Assessments of odour potential were made on three sludges after various storage times: untreated sludge, lime treated sludge, and sludge adjusted to the same pH as the lime treated sludge, but using sodiun hydroxide.
through tubing and fittings made of PTFE. Analysis was undertaken by the Warren Spring Laboratory of the Department of Trade and Industry, according to the method described by Bailey and Bedbo rough The results are shown in Table IV. and plotted in Fig. 3. and 4. Table IV. Variation of odour strength of extracted samples with volune of eluted air Volume of air Strength of odour samples passing through (dilutions) sludge before sampling (1/1) Raw sludge Digested sludge 0 154 000 9 900 11.1 53 000 350 22.2 30 600 270 55.6 15 500 190 111 8 200 160 It is clear from these results that there is considerable die-off of odour strength with time, and that, as would be expected, the anaerobic digestion of sludge can reduce the odour potential by at least one order of magnitude. To illustrate the importance of this die-off effect, the results have been re-plotted in Fig. 5. in a cunulative form; that is to say as cumulative percentage of the eventual colour release against volume of air. In the case of the raw sewage sludge, 38% of the ultimate odour was carried in the first odour sample, and 90% of the odour had been extracted by the passage of about 200 1. In the case of the anaerobically digested sludge, the same effect is much more marked; 72% of the ultimate odour was carried by the first sample, and thereafter the strength of the odour fell off very rapidly. There are two possible explanations for this. First, it can be postulated that as it is known that many of the important odorous chemical species are highly volatile, they may be only physically trapped in the sludge, and need little encouragement to transfer to the atmosphere. An alternative explanation concerns the existence of two equilibria. As the vapour/liquid equilibrium is disturbed by the passage of air, the concentration of dissolved compounds in the liquid phase falls, disturbing the ’solid’/liquid equilibrium The kinetics of transfer across this latter phase boundary are much slower than for the liquid/vapour transfer, so that the extraction of odour becomes limited by the rate of diffusion into the liquid phase. Two observations may be cited as evidence for this latter view. First, when sludge is applied to land, there is a rapid tail-off of odour nuisance after spreading. Hie incidence of rain after a dry period is known to result in an increased evolution of odour. Second, in earlier experiments samples of sludge were centrifuged, and the supernatant liquor discarded and replaced by tap water, before being used in the standard odour potential test. Some re-extraction of odour from the samples was rapidly found. In practice, both postulated mechanisms are probably at work, especially if the concept of ’solid/liquid equilibrium’ be extended to
encompass equilibria governing the production of odorous compounds by chemical and biochemical reduction. 4.5. Wayj5_IrL._Vihich The 'Odour Capacity' Test Might Be Used To Predict Odour .Nuisance Fean .Spreading There are several conclusions of practical importance to be drawn. First, it is clear that the EEC sludge-to-land which directs that sewage sludge spread onto agricultural land should be "stabilised" has beneficial effects on the mitigation of odour nuisance, since both the strength and eventual total evolution of odour is cut by a factor of ten by the principal UK method of stabilisation. Second, the most important potential sources of nuisance in the disposal of sludges to land are the transfer of liquid sludge to a tanker, and the act of spreading itself. This is especially so with anaerobically digested sludges. Third, it seems likely that this source of nuisance could be considerably ameliorated by the simple precaution of aerating the sludge before disposal. Work has shown that the aeration of sludge
different aspects of slurry disposal, for example, spreading, and odour from land after spreading. Correlation work to establish which fractions are best remains to be done, and will probably involve chemical analysis. It is clear then, that the concept of Odour Potential of a sludge or slurry can be a useful tool not only to the odour specialist but also to the operators of sewage treatment works and agricultural slurry handling facilities in minimising the nuisance from the disposal of sludges. (1 j GILLARD, F. ’Measurements of odours by dynamic olfactometry; application to the steel and carbonisation industries." Paper presented at internation symposium: Characterisation and control of odoriferous pollutants in process industries, Belgian Filtration Society, Louvain-La-Neuve, April 1984. (2) TOOGOOD, S. J., and HOBSON, J.A. ’The Discharge of Volatile M rials to Sewers", Water Research Centre, Technical Report TR142, England, 1979. (5) Commission of European Communities, ’The use of Sewage Sludge in Agriculture’, Draft Directive, November 1981. (6) HURLEY, B. J., and RACHWAL, A. J. "Reducing Sludge Volune" Effluent and Water Treatment Journal, 21» 292-296, 1981.
(3) TAYLOR, P. L. "Odour Nuisance from Sewage Sludge", Ph.D. Thesis, University of Bradford, 1982. (4) BAILEY, J. C. and BEDBOROUGH, D. R. "Sensory Measurement and Instrumental Analysis of Odours in ’Septic’ Sewage - Problems and Solutions" Institute of Water Pollution Control, Maidstone, Kent,
1. not annoying 2. a little bit annoying
COMPARISON OF OLFACTOMETRIC ODOUR MEASUREMENT AND CHEMICAL ODOUR MEASUREMENT N. SCHAMP and H. VAN LANGENHDVE Laboratory of Organic Chemistry, Faculty of Agricultural Sciences State University of Ghent, Belgium Summary
Chemical analysis of odorants in ambient air is hampered by the presence of a plethora of volatile organic compounds, which do not contribute to the odour. Nevertheless application of either powerful separation and identi fication techniques, such as the GC-MS combination, or specific GC-detec- tion or absorption procedures allow qualitative and quantitative determi
concentrations of odorcxis carpaunds with annoyance have still to be impro ved. At this mcment the knowledge in this field is far inferior to the knowledge for the sense of view and the sense of hearing. But there is no reason why we would not acquire the same understanding. Although knowledge on the correlation of odorous compounds concen tration and odour impression is still limited, it is used in all types of olfactometry. Indeed diluting this concentration by adding pure air is a general practice. Also many investigations were performed where che micals are added to air and used in psychophysical experiments. Many spea kers in this workshop will present data in this field. Here only chemical analysis will be dealt with. 2. PRINCIPES The goal of chemical analysis of odorous compounds in air is to de termine all substances, which interact with odour perception cells in our nose, both qualitatively and quantitatively. However, with a few excep tions all carpcunds with certain vapor pressure have an odour, meaning that their volatilized molecules react with the membrane of odour recep tor cells. As will be shewn, always hundreds of compounds are present in air; this means that the analysis would be very complex. However as was said before, our sense of smell is selective : for some products it is very sensitive for other compounds it is much less sensitive. Table I : Odour threshold values (ppb) of some organics Compounds Odour threshold Compound Odour threshold butane 1.3 106 acetic acid 40 butane 500 propionic acid 190 butanol 300 butyric acid 38 butanal 15 valeric acid 8 butanethiol 0.8 hexanoic acid 42 Several extensive lists of threshold values, i.e. the minimum concentration in air, that is detected by 50% of the population, have been published (1, 2,3,4). However published threshold values for a particular compound can vary over a number of orders of magnitude, so they have to be treated with scepticism. This selectivity makes chemical analysis of odour easier : many com pounds, although present in ambient air, and although they have an odour in pure form, are not contributing to the odour, while their concentration is far inferior to the threshold value. On the other hand the sensitivity is high for a range of compounds, higher than any chemical analysis can cope with directly. These canpaunds have to be concentrated frcm the odorous air, so that higher amounts are available for the analytical technique. If this concentrating could be done with the same selectivity of odour recep tor cells, there would not be much of a problem. Hcwever the actual know ledge of this interaction is far too limited - in fact it is inexisting -to speculate on an analytical application. With all of the biochemical de velopments, it is not excluded that at a certain mcment it beccmes feasible, but right new the only way is to use crude physicochemical methods, such as
freezing out, adsorption and absorption. After concentrating, separation is achieved by classical methods such as gas chranatography (GC) or high pressure liquid chranatography (HPLC). Identification is based mainly on mass spectrometry, infra-red spectrometry and chrcmatographic data. 3. RESULTS The primary goal of these methods is to concentrate all volatile com pounds, mainly volatile organic compounds or VOCs, present. This mixture of VOCs, containing odorous ccmpcunds, next to a large majority of unodo-rous substance, then is analysed. This chemical analysis is based on the separation of these hundreds of compounds by gas chranatography, is hampe red by large amounts of water, which is always present in air, and which is also freezed out or adsorbed. The only way to escape more or less this difficulty is to use a rather apolar adsorbant, in casu Tenax GC or similar materials (e.g. Chranosorb 102) (5). A second limitation is the fact that no material will ever be capable of adsorbing all odorous com pounds completely, and permit to desorb then afterwards completely. For compounds with very low boiling point, e.g. hydrogen sulphide, strong ad-sorbants are necessary, while for odorants with high boiling point, e.g. skatol or the sesquiterpenes, thermal desorption is difficult with strong adsorbant s. So a compromise has to be accepted, or several complementa ry adsorbants have to be used. At this moment this compromise for concen trating all odorous substances is found in the adsorbant mentioned, kno wing that the most volatile compounds might escape partly. Many systems have been described and even carenercialised, but we use a home-built sy stem, which is schematically represented in figure 1 (6). On an outer side wall of the gas chromatograph (GC) an oven in which the Tenax-adsorp-tion-sampling tubes fit is constructed. Connections with pressurized he lium (transfer gas) is provided and their is a connection with a high tem perature resistant sixway valve, which replaces the normal GC-injector. During thermal desorption (position 1 in figure 1) the transfer gas, car rying desorbed volatiles, passes the sixway valve, a cold trap (stainless steel loop cold with liquid air) and enters the ambient air. The helium carrier gas is connected to the GC-column via the sixway valve. After the desorption stage which usually takes about 45 minutes, with a desorption oven temperature of 220°C for 30 minutes at least, the sixway valve is switched (position 2 in figure 1). At that moment transfer gas flows through the sixway valve directly into the ambient whereas the carrier gas passes the cold trap before entering the GC-column. The liquid air is removed from the cold trap and the latter is quickly heated by a high in tensity fload light. In this way condensed compounds are flash-evaporated and injected into the GC-system. Concentrating odorants by adsorption-desorption techniques produces a terribly complex mixture of VOCs, which is separated by gas chranato graphy. Fortunately this technique allows formidable separation power, but still then the result is not always sufficient far a clear-cut odour analysis. In figure 2 the GC-analysis is shown of an air sample in the neighbourhood of a rendering plant, showing a great number of VOCs; however almost all of them are hydrocarbons produced by cars and heating systems and sane other products, which do not contribute to the odour. Very small peaks of odorants are detected, which shows the difficult task of odour ana lysis with a general concentrating technique. Of course this analysis is far more relevant if emission gases are examined as is demonstrated in fi gure 3 (7). Part of these difficulties can be overcane if the odorants can
be detected specifically, which is possible for sane groups of odorants (thiols or mercaptans, sulphides, amines) with specific GC-detectors. Spe cific detectors are available for haloganted compounds, sulphur-, phosphor-and nitrogen compounds. Figure 4 shews the analysis of the sulphur-ccmpounds produced by the acidic decomposition of phosphate-rock and causing the typi cal smell of fertilizer plants. Another approach is to aim at selective concentration methods. Indeed odour problems are caused by a limited number of compounds, on rather a li mited number of classes of compounds, mentioned in figure 5. For most odour nuisance problems, chemical plants, refineries, live stock production, food processing, rendering, water purification plants etc., the compounds responsible for the odour are known. So chemical analysis of the odour can be limited to these odorants, and selective concentrating techniques can be used. Selective concentrating methods are based on speci fic absorption techniques, using particular chemical reactions of odorant classes. Semet imes several absorption methods have to be used in order to describe the odour problem, thus increasing the labor cost of the analysis. On the other hand absorption methods allow better quantitative results. Se lective absorption of odorants from air produces a far less complex mixture. We developed or are developing several of these methods for aldehydes, amines, acids, thiols etc. Carbonyl ccnpounds for instance can be trapped by absorption in a rea gent solution containing 2,4-dinitrcphenylhydrazine and hydrogen chloride. Details of this method are extensively described elsewhere (8). The prin ciple of the method is that the carbonyl ccnpounds, in case of rendering plant emission the aldehydes, react with the 2,4-dinitrophenylhydrazine and form 2,4-dinitrophenylhydrazones (2,4-DNPH's) according to the scheme. These 2,4-dinitrophenylhydrazones have seme interesting properties. It are cristalline caipounds so that after extract of the 2,4-DNPH's fran the reagens, they can be concentrated by evaporation of the solvent without losing product. Besides these caipounds shown intense absorption of UV-light (X 356 nm) and so they can easily be detected with an UV-detec-tor. These properties make the 2,4-DNPH's particularly suitable for HPDC-analyse. This methods is used since seme time. A chranatogram is given in figure 6 and results of the quantitative determination of carbonyl com pounds in different situations are given in table 2. For amines absorption in an acid solution, or preferably adsorption onto an acid ion exchange column (acidified divinylbenzene-styrenesulfo-nic acid copolymer) is used. 10-50 1 of ambient air is sent over*a wet 100nnix3irmI.D. column; the ion exchange polymer is put into a vial, made alkaline and the water solution is analysed on packed Carbowax-KDH GC-column with a thermionic selective detector (TSD), which is specific for nitrogen- and phosphorus-catpounds. Trimethylamine is detected easi ly at 1 ppb. Aibids can be absorbed specifically in an alkaline impringer, which is extracted with ether after acidification to pH 2. This method was used for rendering plant emissions, shewing a series of linear and branched
Table II : Quantitative determination of carbonyl compounds at different odour sources (concentrations in ppb) Rendering plant Gelatine plant neighbourhood neighbourhood Formaldehyde 40 16 Acetaldehyde 39 24 Acetone 36 73 Prcpanal 10 -Isobutyraldehyde 10 30 Pentanal 15 19 Hexanal 3.52 Heptanal 12.5 Octanal 10.5 Nonanal 1 2 acids (figure 7). However extractions always involve a serious decrease in sensitivity, while evaporation of the extract produces a solution in 0.1-0.5 ml of solvent, and only 1 pi of it can be brought in the gas chromatograph. Therefore work is in progress to enhance sensitivity by converting acids in to halogenated derivatives, which can be GC-analysed with the more sensitive electron-capture detector. For thiols a similar procedure is investigated as with aldehydes. One possibility is absorption of thiols in an alkaline solution and reaction with 2,4-dinitrochlorobenzene, yielding 2,4-dinitrofenylsulfides, which are analysed by HPLC (9). Sane improvements on removal of reagents at the one hand and on separation of sane by-products on the other hand have to be achieved in order to in crease the sensitivity with another factor of ten. 5. CONCLUSION The actual scope and limitations of chemical analysis of odour show that all problems can be tackled as far as emission is concerned. For iititiission measurements seme progress is necessary, but there is no essential reason why chemical analysis would be unable to attain the desired sensitivity for all types of odorants. There is no doubt that in a few years the last dif ficulties will be solved. In order to achieve real control of odour nui sance, automatic measurement is necessary on a long time basis. There again seme technical development is to be expected. Does this mean that machines are going to decide if an odour is pre sent or not? By no means, while the population will always be the reference, and psychophysical measurements will be necessary to make chemical analysis possible.
REFERENCES 1. Compilation of Odor and Taste Threshold Values Data, Ed. F.A. Fazzalari, ASTM Data Series DS 48A, American Society for Testing and Materials, Philadelphia, 1978. 2. Compilation of Odour Threshold Values in Air and Water, Ed. L.J. Van Gemert, A.H. Nettenbreijer, RID, Voorburg, CIVO Zeist, Netherlands, 1977. 3. F. Patte, M. Etcheto, P. Laffort, Selected and Standardized Values of Suprathreshold Odor Intensities for 110 Substances, Chemical Senses and Flavour, 1, 1975, 283-305. 4. P.H. Punter, Measurement of Human Olfactory Threshold for several Groups of Structural related Compounds, Chemical Senses, 7(3/4), 1983, 215-235. 5. N.P. Cemansky, Diesel Exhaust Odor and Irritants : A Review, J. Air Pollut. Control Assoc., 33(2), 1983, 97-104. 6. H. Van Langenhove, N. Schamp, Chemical and Olfactometric Measurement of Odours, In : "Characterization and Control of Odoriferous Pollutants in Process Industries", Ed. Societe Beige de Filtration, Louvain-la-Neuve, Belgium, 1984. 7. H. Van Langenhove, F. Van Wassenhove, J. Coppin, M. Van Acker, N. Schamp, GC-MS Identification of Organic Volatiles Contributing to Rendering Odors, Environ. Sci. Technol., 16(12), 1982, 883-886. 8. H. Van Langenhove, M. Van Acker, N. Schamp, Quantitative Determination of Carbonyl Ccnpounds in Rendering Emissions by RP-HPIC of the 2,4-dinitro-phenylhydrazones, The Analyst (London), 108, 1983, 329-334. 9. H. Van Langenhove, M. Van Acker, H. Van Langenhove, Separation and Deter mination of 2,4-dinitrophenyl thioethers by RP-HPLC, J. Chranatogr. 257, 1983, 170-173.
FIGURE3 _ -C HROMATOGRAMOFTHEGC -M SANALYSISOFARENDERINGPLANTEMISSIONSAMPLE , ODORANTSIDENTIFIED : ° 1 . trimethylamine15 . pentanal2 . dimethylsulfide16 . dimethyldisulfide6 . propanal19 . hexanal7 . 2 -m ethylpropanal25 . heptanal8 . 3 -m ethylbutanal26 . dimethyltrisulfide9 . 2 -m ethylbutanal37 . nonanal
FIGURE 4 Chromatogram of the Tenax-adsorption FPD-analysis of volatiles emitted during acidic decomposition of phosphate rock. Compounds : 1. dimethyl sulfide 2. dimethyl disulfide 3. methyl isopropyl, disulfide 4. alkyl substituted thiacyclopentanes and alkyl substituted thiacyclohexanes.
Oxygen compounds Aldehydes RCHO R = CnH2n+1 - CnH2n-l Carboxylic Acids C H~ ,.COOH n = 1 to 6 OH n 2n+1 Phenols I °'T ~ ’R R = H or C_H
n 2n+l 2. Nitrogen compounds Amines R-.R-R^N R = H or C H_ Indols I j R = H
USE OF PEAT AS LITTER FOR MILKING COWS I* Peltola Work Efficiency Association, Finland This study compared the advandages and disadvandages of peat, straw and sawdust for use as litter. The way in which peat is used, the amounts used and the effect of the litter on the structural parts of the tying-stall shed and an the labour requirement were investigated, and the quality of milk, the incidents of mastitis and the value of the manure were studied. The results show that peat absorbs urine and binds airmonia better than the other litters tested. Peat manure contains more than the average amounts of nitrogen and magnesium, and the nutriens are in the form that is more readily utilised by plants. The anmonia contents of the cowshed air were slightly lower with peat litter than with either straw or sawdust. The difficulty in peat was in handling it. There were no significant differences between the three litters in terms of the labour required. On the other hand, peat was more difficult to store during the cold winter because it tended to freeze. Urine separation systems were easily blocked by peat. The dust content of the cowshed air rose when peat was used. The litter had no affect on the state of health of the animals or on the quality of the milk. These factors are affected more by conditions on the farm in question. Peat was found to be suitable for use as litter. Flexible use of peat requires storage, spreading method and manure removal be designed specifically for peat. These factors are being studied in the final part of the work, which is still in progress. 1. INTRODUCTION The aim of this three year joint study was to investigate the advantages and disadvantages of peat, straw and sawdust for litter as comprehensively as possible. In Finland straw is used on 67% of farms, sawdust on 25% and peat on 3% of dairy farms. Only about 3% of farms use no litter at all. Itie use of peat as litter was compared with sawdust and straw on 15 dairy farms during the indoor feeding period of 1983- 84. For the first 3 months 5 farms used straw as litter, 5 used sawdust and 5 peat. At the end of this period all the farms changed over to peat litter. All the cowsheds had tying-stalls, from which the manure was removed in solid form. The use of peat as litter was studied in Finland in the 1930s and 40s. Today, peat is harvested mechanically by means of a milling cutter, which creates fine particle peat. Cowsheds have also changed with respect to the use of peat with the mechanisation of manure removal. At the same time workers are now demanding better working conditions.
* PROPERTIES of litter peat Litter peat is the surface peat removed from the bog before cutting of fuel peat. As fuel peat production increases, so the amount surface peat removed also grows. Around 1 million m3 of surface peat is lifted every year in Finland, and the figure is expected to rise to 2-3 million m3 a year by the 1990 (1). Surface peat is used mainly for horticultural purposes. The properties of the peat vary, depending on the place in the bog frcm were it canes. The surface layer provides the best litter peat. Lifting peat frcm deeper in the bog results in the inclusion of layers of more highly decomposed peat, which is not as good for use as litter. Freshly cut peat usually contains 40-60% water. Litter peat, however, must not contain more than 40% water, since its absorption capacity and storability both decrease with increasing water content. Good quality litter peat should be Sphagnum fuscum peat. A suitable degree of decomposition is 2-3. Raw Sphagnum peat is acidic, with a pH of 3.0-4.5. Its total nitrogen content 1.0-1.5% of dry matter. 3. AMMONIA-BINDING CAPACITY Because its acidic character, peat binds ammonia well. The amnonia-binding capacities of the different litters were measured in the laboratory. Aqueous ammonia solution was applied to the peat, which was then dried under reduced pressure. The results showed that peat can bind up to 2.5% of its dry weight of ammonia (Fig. 1). The anroonia is so strongly bound that it does not evaporate even when the peat dries. The binding capacity of straw and sawdust is less than 1% (2). From the point of view of the binding of ammonia, it is important that the peat is Sphagnum fuscum peat. Earlier studies (3) have shown that other varieties of peat bind only 0.26-0.86% of their dry weight of ammonia.
Figure 1. Anmonia-binding capacity of the three litters under laboratory conditions. 4. LIQUID-BINDING CAPACITY The absorption capacity of the litter determines the amount used, and also affects the functioning of hydraulic manure presses. Hie absorption capacity of litter depends on its initial moisture content. In the laboratory tests , peat had a much greater absorption capacity than other litters. Peat absorbed a maximum of 4.5 times its own weight of liquid, straw 3.5 times, cutter shavings 3.6 times and sawdust 1.5 times (4). The litters used in the tests had an initial moisture content of 20%, with the exception of peat, which contained 40% moisture. These moisture contents represent the working values for good-quality litters. Peat was far superior in terms of absoption capacity to the other litters. Litter should retain its absorption capacity even when under pressure, for example in a hydraulic manure press. The ability of litters to withstand pressure was tested by first allowing than to absorb either water or urine. The litters were then compressed in a hydraulic press for one hour as the pressure was increased stepwise from 20 N/an2 to 50 N/an2. The pressure exerted by hydraulic manure presses in practice is around 40-60 N/cm2. All the retained urine better than water. Sawdust was found to have the greatest resistance to pressure, retaining about 75% of the absorbed
liquid. Cutter shavings retained 40% of water and 52% of the urine, while chopped straw retained 46-50% of its absorbed liquid. Peat showed the lowest retention of liquid, keeping only 19% of the water and 33% of the urine (Fig. 2). Figure 2. Effect of pressure on water-binding capacity of litters. The results indicate that special attention should be given to the efficiency of urine separation on farms using peat. Manure presses should be fitted with some means of urine separation so that urine pressed frcm the litter can be led out of the press chute. This is essential for the smooth operation of the manure press. The urine separation system in the troughs, perforated plates or scale plates may become blocked when peat is used. Provision should be made for flushing out the urine separation pipes in the case of a blockage. Various types of perforated plate are currently being studied with respect to their susceptibility to blocking. 5. NUTRIENT-BINDING CAPACITY Samples of manure were collected from the farms during both experimental periods. The samples were used for nutrient content determinations and for a pot experiment in which Italian rye grass was cultivated. The nutrient contents obtained were compared with the
values far the farm in question during the use of different litters, and with the grouped means for the different litter manures. Peat manures contained statistically significantly more total nitrogen, magnesium and dry matter. The lowest nutrient contents were found in the sawdust manures. The pot experiment showed the same differences. The rye grass was cut four times. The greatest differences in the increase in dry matter yield were obtained with the second crop. As all pots received the same basic phosphorus, potassium and magnesium fertilizer, the differences in yield were due to the amount of available nitrogen contained in the litter manures (Fig. 3). In addition 31% of the total nitrogen in the peat manure was comparable with the nitrogen present in ccnmercial fertilizer, canpared with only 19% in the case of straw manure and sawdust manure (2). o O)
over a period of 1-3 hours. Dust samples were collected by drawing the air through a filter at the rate of 1.71/min. The filter pore size was 0.8 um. The dust contents of the air were higher when peat was used than when the other litters were used (5). Table I: Dust contents of air during use of different litters. Fixed measuring point. litter dust content of air standard no. of mg/m3 of air deviation measurements peat alone 1.24 0.65 11 peat and straw 0.53 0.29 6 straw alone 0.10 0.20 4 straw and sawdust 0.24 0.28 4 sawdust alone 0.13 0.23 3 all farms 0.67 0.66 28 Variance analysis showed the diffemces between the groups to be highly significant F(4,23) = 7.43, p< 0.001. According to the t-tests, there was a significant difference between peat alone and all the other litters, and also between straw and straw and peat combined. 6.2. NH3 CONTENT OF AIR --------------- 3 The differences found in the NH content of the ocwshed air during the use of the different litters were not as clear as in the case of dust contents. The contents measured were generally very lew. On farms using peat, the aimonia content of the at the tying-stall at milking height was 2.1 ppm, whereas on the other farms it averaged 3.2 ppm. The corresponding ammonia contents during manure removal were 3.5 ppm for peat and 4.0 ppm on the other farms (5). The aimonia content of the air was determined using a Drager detector tube, which is not very accurate, particularly at low concentrations. In theory the acidic nature of of peat suggests it should bind armonia far better than the other litters. Hcwever, the results obtained indicate that the use of peat freshens up the cowshed air only if the manure removal and urine separation systems are working efficiently. On the basis of a sensory evaluation, the air in cowsheds using peat was slightly fresher than that in the other ccwsheds. The same result was obtained in the farm interviews. This study revealed no obstacles to the use of peat as cowshed litter. The quality of the milk and the health of the animals are affected more by the general standards of hygiene than the type of
litter used. The advandages derived fran using peat depend greatly on the qualiity of the peat. The absorption capacity of dry peat is best exploited in cowsheds were all the urine is absorbed by the litter. In this case, the amount of peat required is far less than that of either straw or sawdust. In cowsheds with a urine well, the use of peat as litter permits some reduction in the size of the well. The economic advandage of using peat lies in the better utilization of nutrients. At 1984 prices, the saving achieved by retaining the nutrients is about 6 Finnish marks (about 0.9 USD), per cubic meter of peat used. The price of peat liitter is about 20 marks/m3 (3 USD). Additional benefit is derived frcm the soil improvement effect resulting from long-term use, and frcm the lower cost of constructing the building if the urine well can be made smaller or emitted altogether. References 1. KYTO, M., SIPILA, K. & THUN, R. 1983. Pintaturve maanparannusaineena ja kuivikkeena. Sumnary: Weakly decomposed peat for soil improvement and litter. Technical Research Centre of Finland. Research notes 240. 2. KEMPPAINEN, E. 1984. Kuivikkeen vaikutus lannan arvoon. Effect of litter on value of manure (in Finnish). Kasikir-joitus. Manuscript. 15 p. 3. TUQRILA, P. 1929. Bindungsvermogen verschiedener Torfarten fur Stickstoff in Form von Ammoniak. Ability of different peat types to bind nitrogen in the form of ammonia (in German). Sucmen Suovilj. yhd. julk. 9. 4. PELTOLA, I. 1984. Kuivikkeiden nesteenpidatyskyvyt testissa. Summary: The absorptivity of different litters. Tyotehoseu-ran rakennustiedotus 200. 6 p. 5. PELTOLA, I. 1984. Kuivikkeen vaikutus navettailmaan. Summary: The effect of dry litter upon tying-stall shed air. Tyoteho-seuran maataloustiedotus 317. 4 p.
piggeries increased. In that time limits to farm size were seldom determined by farm waste disposal and air polution. At present there are local regions where is a surplus of slurry. Due to the low value as a fertilizer transport to other regions is too expensive and in some cases over manuring will exceed. Separation of slurry offers the possibility of obtaining solid manure with a high fertilizer value and a liquid with a low value. On this way it becomes more worthwhile to transport the solid manure over longer distances, So it reduces over manuring. However slurry separation with a high efficiency is costly about £4,-/pigplace by adding flocculants and expensive complicated separations (1). Underslat slurry storage also influences environnement as well inside as outside the piggery because gases are released from the slurry. On the other hand anaerobic digestion processes will be avoid by daily removal and separate storage of faeces and urine. That factors were for IMAG a motive to develop a reliable combined manure filter and removal system for piggeries. 2. SEPARATION EFFICIENCY In a preliminary investigation the separation efficiency of different separation techniques under slatted floors was defined. Separation efficiency means which part of the total components in faeces and urine remains in the faeces. This research was carried out in a pen with 8 pigs. The pigs were given wet feed without drinkwaterprovisions. In Fig.l the separation result is mentioned for a filter net of meshsize 0.78x0.78 mm. From Fig.l it is apparent that about 35% of the total faeces and urine is removed as a solid and that about 90% of the total dry matter is in the faeces. Also for a number of minerals, p2°5’ Ca(^ » and Cu> amounts to more than 90%. Nitrogen and potassium were separated in smaller amounts, about 60% and 40% respectively, being retained in the solid. On basis of this result and after comparative research with concern for filtering, clogging and cleaning the above mentioned filternet is now being used for the final mechanized filter and removal system. 3* THE MECHANIZED FILTER INSTALLATION A combined filter and manure system must be completely reliable, since it is not easy to make repairs under “the slats.v After 2-3 years of experience a system which works well has been developed.(2,3,4) In Fig.2 a schematic diagram of the filter system is shown under the slatted floor. In the channel under the slats two angle sections (1) are attached one above the other and fixed over the whole length to both walls. These are covered from above with protective plates (2) joined underneath the slats. The filter net (3) is provided with
Table 1: Average composition of faeces and urine after separation by a filter net (0.78x0.78) Faeces Urine Dry matter (%) 32.50 1.92 Crude ash of dry matter 25.70 63.10 N-Kj (%) 1.24 0.34 NH -N 34 0.35 (%) 1.64 0.05 K20 85 0.62 CaO 45 0.04 MgO 48 0.02 Cu ppm 197 2.50 pH 9.1 From Table 1 it is evident that the percentages of the minerals in the faeces are high. In the urine the percentages are much lower except potassium. By manuring grassland potassium is the limiting factor, while o.n arable land the quantity of nitrogen needs to be taken into account. In applying solid manure to cropland the Cu-content needs to be taken into account. Depending on the Cu-status of the soil, 0-6 kg Cu/ha is advised. By fertilizing with 10 t/ha of solid manure about 3 kg Cu/ha is administered Because only a small amount of copper is taken up by plant growth and lost through drainage, the application of solid manure needs to be spread out over a few years if Cu is not to accumulate in the soil. 5.2 Odour_emissions It can be concluded that separation and removal of urine and faeces from piggery result in a reduced formation of odour components (5). This might result in a decrease of the precieved odour as compared with a housing system with underslat slurry storage. In order to obtain a reliable figure for the actual odour reduction, measurements have been carried out. Samples of ventilation air from a pighouse with underslat slurry storage as well as a pighouse with filter nets were taken on a number of different occasions. All samples were collected in bags made from FEP-Teflon. Odour experiments were performed the following day using a dilution apparatus (olfactometer) and a group of observers (panel). Since the establishment of the odour intensity is a time consuming affair, it has become practice in Dutch agricultural odour research to concentrate on the establishment of the odour treshold (6). The odour treshold is defined as that dilution of odorous air which
cannot be distinguished from odourless air by 50% of the panel members (DT50). This implies that the threshold is a barely detectable odour. The number of times a sample has to be diluted to reach threshold levels is a measure for the relative strength of the odour. The relative odour strength times the ventilation rate of the building results in the odour emission. This can be regarded as the total odour load per unit of time leaving the building. Finally the odour emission can be used in atmospheric dispersion models in order to calculate the odour threshold distance. Table 2 shows the results of the experiments as well as the relevant data of the pighouses at the time of sampling. During the measurements the ventilation rate between the pighouses varied. The difference are due to different ventilation rates and due to sampling in the morning or in the afternoon at different ambient temperatures. Table 2: Odour measurements Pighouse with separation Data of sampling 24.5.83 31.5.83 14.9.83 28.10.83 Number of pigs 158 158 158 157 Average liveweight (kg|1 _1 75 80 45 75 Ventilation rate (m3kg~ h~ ) 0.61 0.93 0.89 0.47 Dilutions to threshold (DT50) 770 1008 817 1634 Total odour emission (DT50/h.103) 5595 11902 5195 9103 Odour emission/pig (DT50/h) 35410 75326 32877 57980 Emission reduction/pig (%) 49 50 50 59 Pighouse with underslat slurry storage Data of sampling 24.5.83 31.5.83 14.9.83 28.10.83 Number of pigs 300 275 279 279 Average liveweight (kg) 80 90 45 85 Ventilation rate (m3kg~ h- ) 0.21 0 54 0.52 0.57 Dilutions to threshold (DT50) 4133 3068 2820 2903 Total odour emission (DT50/h.103) 20632 41234 18409 39205 Odour emission/pig (DT50/h) 68773 149942 65982 140520 Emission reduction/pig (%) n.a. n.a. n.a. n.a. n.a.= not applicable It can be concluded from Table 2 that the installation of filter nets reduced the odour emission per pig by approximately 50%
ECONOMIC ASPECTS 6.1. In Fig. 3 the lay-out of the piggery is mentioned. The stall is renewed and instead of the normal depth of 1.50 m by underslat slurry storage the channels have a depth of 0.50 m. These shallow channels are as a matter of course cheaper but on the other hand facilities for storage of dung and liquid must be build outside the building. In front of the stall a dung and liquid gutter has been build, where the transportbelts are installed. The investments for the filter system of this piggery are given in Table 3. Table 3: Survey of extra capital investment of the filter system, in comparison with the slurry system (£/pigplace) Filter system in both channels 3750 Dung and liquid gutter 1250 Transportbelts 2500 Dung and liquid storage 2500 10000 Estimated reduction by shallow channels 2500 Extra capital investment 7500 Extra capital investment/pigplace 46.90 This price is relatively high on account of the small number of pigs. Other calculations show that the investment costs are lower in comparable stalls with a capacity for 480 and 960 pigs. These stalls are more common, the costs are respectively £ 18 and £ 11/pigplace. Calculations confirm also the expectations that stalls with long channels in longitudinal design are more favourable for this filter system than stalls with short channels in transverse style. 6.2. Annual costs The costs depend first on the capital costs and secondly on the costs of straw. On the other hand savings are possible on building structures and heating system. Besides these factors other savings are possible because some disadvantages of the slurry system can be removed by the filter system; for example costs for fuelexpenditure and slurry disposal. A calculation for a standard plan for 480 and 960 pigs shows that in a comparable stall as in Fig. 3 the extra annual costs for the filter system are slightly higher than for the slurry system; respectively £ 1.11 and £ 0.08 /pigplace. The costs for stalls with transverse and short channels are much higher and vary between £ 3.70 - £ 3.40 /pigplace. Compared with the costs for slurry separation in a slurry disposal centre the costs for the filter system in stalls with long
G.-O. MEJER Institut flir landtechnische Grundlagenforschung der Bundesforschungsanstal t fiir Landwirtschaft Summary As the raw material of rendering plants produces very odoriferous
/3-5/, a series of olfactometric measurements in working rendering plants were carried out. Furthermore, the results should help to prevent mis-planing and misinvestments due to misunderstanding odour limit values. 2. EXPERIMENTAL 2.1 RENDERING PLANTS AND AIR CLEANING SYSTEMS In Germany approximately 80 rendering plants, spread all over the land, process more than 800 000 t of raw material per year. Five plants in northern Germany were chosen with different size and different air cleaning systems, table 1. These chosen air cleaners are of the common types in rendering plants. The attempt to include also a bioscrubber and an activated carbon filter failed due to frequent breakdowns. All air cleaners worked already for several years and were operated in usual manner without any variations or make up especially for the measure ments. These were carried out in a moderate weather period with ambient temperatures of 16-21 °C. The samples were taken during normal continuous plant performance, that means with closed cooker. In addition, in plant A samples were taken in peak load performance, when the cooker was opened. Plant ABCDE raw material animals animals bones animals blood /parts /parts /parts capacity t/h5 12 10 distance to m next housing 900 1500 800 air cleaning counter cross lime bio bio system current current stone filter filter scrub scrub tower ber ber reagents 2 , 2 , 2 , carbage carbage compost compost H2°2 24, 2, 2 2
In the vicinity of a rendering plant it is very difficult to find a really unpolluted place for the olfactometric measurements. Although the most unpolluted site luff of the plant was chosen, another precaution was taken. As adaption of the panelists to the plant odour could cause one of the greatest errors, some minutes before and during the measurement the panelists inhale solely odourless air from the olfactometer. To prevent discomfort by inhaling completely dry air, the olfactometer Modell 1158 is supplied with a moistening device, fig. 1. Fig. 1. Moistening device. In a standard impinger, filled with destillated water, air is moistened close to saturation. An equal flow of moistened air is mixed to the olfactometer outlet, thus delivering to the panelist a rel. moisture content of nearly 50 %. The panel consisted of 4 persons. The samples are prediluted taken into plastic bags, simultaneously at the inlet (raw air) and at the outlet (cleaned air) of the air cleaners. To receive an unfalsified sample from the outlet of the biofilters, undiluted by ambient air, a "tent” of plastic foil, fig. 2, is placed on the filter surface. The cleaned air blows up the tent and escapes through the sample hole, large enough to prevent a significant increase of pressure. The form of the upblown tent indicates, wether a sample area with normal air flow is chosen, and over the space of the covered filter area of 6,25m2 an average sample is received. Fig. 2. Device for cleaned air samples from biofilter outlet.
RESULTS AND DISCUSSION The olfactometer readings of the measurements are statistically treated as described in /3/. The results for the plants and air cleaning systems, described in table 1, are given in table 2. system chemical biological plant ABCDE rel. odour raw air 65200 14200 26800 41400 95100 concentration Z50 cleaned air 48300 7360 29500 7930 5100 /odour units/ olfactometric efficiency n 26 % 48 % 81 % 95 % Table 2. Results of measurements, obtained during normal performance, cooker closed. Taking the index R for raw air at the cleaner inlet and the index C for cleaned air at the cleaner outlet, the olfactometric efficiency of the cleaner is defined according to /6/: 50 R In the regarded air cleaning systems, the odoriferous pollutants are first seperated from the raw air by sorption and then decomposed by chemicals or by micro-organisms. As long as this decomposition is not yet completed, the pollutants may desorb and repollute the air, when sorption conditions, i.g. the raw gas concentration, change. By the relation of the difference in raw and cleaned gas concentration to the actual raw gas concentration, a negative efficiency may be calcula ted by equation 1, i.g. when a low raw air concentration is preceded by a high one. Table 3 shows peak concentrations and increasing olfactometric efficiency, wherT’in plant A the cooker is opened. rel. odour concentration raw air 627000 Zgg /odour units/ cleaned air 240800 olfactometric efficiency n 62 % Table 3. Results of measurements, obtained in peak load performance when cooker is opened. Although the number of measurements is too small for general assertions, some deductions can be drawn: The results confirm the superiority of the biofilters. And in fact, the number of biofilters in rendering plants increases. Concerning the rel. odour concentration in the cleaned air, a large difference is evident between the presented results and the assertion that a limit value of 100 odour units can be achieved. Two interpretations can be offered:
A rendering plant manager is overcharged by the correct maintenance and the correct control of a chemical scrubber according to the changing conditions of his plant. The pollution is blown into the sky and he has no reliable signal to survey cleaning efficiency. So he sets the chemi cals dosage as recommended and trusts in the guarantee of the manufactu rer of the cleaner. And as he has payed a lot of money for it, he is certain he has done his best. Biofilters adapt themselves, and due to the groundlevel outlet a fai lure is easier perceptable. 2. The limit value mentioned above and likewise guaranteed by manufacturers of air cleaners is based on measuring methods different from those applied here and described in the guidelines /3-5/. Results of measure ments, achieved by different methods are not comparable. The large dif ference of the values is a distinct indication of differences in exi sting methods. 4. CONCLUSION The necessity of a harmonisation and standardisation of the complete method of olfactometric measurements is evident, in order to achieve com parable results. A basic requirement for the establishment of any odour limit value is that such a method is established and generally accepted, and that all measurement results on which an odour limit value is based are also achie ved by exact this method. REFERENCES (1) QUELLMALZ, E., Tierkorperbeseitigung ohne Geruchsbelastigung. VDI-Kolloquium "Minderung von Geruchsstoffemissionen", Wiesbaden, Mai 1981. (2) VDI 2590 Entwurf: Auswurfbegrenzung, Anlagen zur Tierkorperbeseitingug. August 1979. (3) VDI 3881 Blatt 1 Entwurf: Olfaktometrische Technik der Geruchsschwel-lenbestimmung, Grundlagen, November 1983. (4) VDI 3881 Blatt 2 Entwurf: Probenahme f'ur die Geruchsschwellenbestim-mung mit dem Olfaktometer. (5) VDI 3881 3 Blatt Entwurf: Messen der Geruchsschwelle mit den Olfakto-metern Model! 1158 und T04. (6) VDI 3477: Bilogische Abluftreinigung. Biofilter, Dec. 1984. Special thanks to the owners and managers of the plants for the spantaneous permission to carry out the measurements, and likewise to Mr. E. GARRELTS for his competent counsel and and angagement in this work.
FIG.1 INSTANTANEOUS AND AVERAGE CONCENTRATION PROFILES Instantaneous Top View Instantaneous 1 Hour Average of Plume Concentration Concentration Profile at X=x0 Profile at X=x0 FIG. 2 ODOUR CONCENTRATION FREQUENCIES (AS % WITHIN EACH 10° WIND SECTOR) NESWN (a) ALL CONCENTRATION , (b) C>C0/16,(c) C>C0/4, (d) C > C0/ 2 , (C0 = MAX. AXIAL ODOUR STRENGTH ) | <684 5)
Fig 2. Apparatus for the desorption of volatiles from Tenax. A C Splitter valve B Toggle valve D Septum purge valve E G Tenax Reduction union F Cold trap H Tenax tube I Septum holder J Desorption heater KM0.2 mm i.d. transfer line L Septum 00.3 aa i.d. capillary N GC Injector QH1/ el 1i6u ” m low dead volune union tee P0.2 mm i.d capillary Glass insert
Odour quality depends not only on the sensitivity of the human nose but also on the subjectivity of the human language to be able to describe the odour (17). Some chemical characteristics of slurry have been compared to the slurry odour. A relationship between volatile fatty acids (VFA) and odour offensiveness of poultry manure was described by Bell (18). Sobel (19) used a panel of people to assess the odour offensiveness of poultry manure and found a correlation between total solids and offensiveness. Williams (5) found that odour intensity measurements by liquid dilution of piggery slurry were not successful, and concluded that the offensiveness of odours rather than their Intensity was the best indicator of odour nuisance.Odour offensiveness is a supra-threshold assessment and depends on the human sense of smell. Since complaints result from the detection of offensive compounds by the human nose and because of the problems of quantifying all the odour compounds in slurry by analytical methods, it is desirable to use the human nose for odour offensiveness measurements. This presents problems, since there is a large variation of sensitivity to odorants and some people are anosmic & 20). The well being of a person also affects their response to odours (2). Odour offensiveness is a subjective determination and requires as many people as possible to assess an odour and minimise variations in sensitivity. The problems of routine olfactory odour assessments are that they are not only very time consuming, but also very subjective. Spoelstra (21) listed criteria for potentially suitable odour indicators and concluded that VFA and p-cresol were the best odour indicators of piggery slurry. Williams (5) examined the supernatant 5 day biochemical oxygen demand (BOD ), total organic acids (TOA), VFA and
An improved model using supernatant BODe and a similar model using TOA as indicators of potential oaour offensiveness is presented. METHODS Raw slurry was collected and prepared by diluting fresh excreta from fattening pigs with tap water to a standard concentration of total solids (TS) 30 g/1 and chemical oxygen demand (COD) 38 g/1. The methods of excreta collection, the animal diets, and the slurry composition were described previously (22 & 23). Treated Slurry Samples of treated slurry were obtained from laboratory-scale continuous culture reactors (3 & 15 litres) during a series of treatments studying the effects on residual slurry quality of mean treatment time, reaction temperature, dissolved oxygen level and pH value (27). Some were also collected from a 2.4m pilot plant which was operating at 35°C and 7 day residence time and with dissolved oxygen saturation of 0 to 40%. The pilot plant was treating separated stored piggery slurry (TS 21 g/1; COD 26 g/1). Measurement __of Biochemical Oxygen Demand (BOD)
The Slurry Supernatant BOD^ was measured by the standard method of the American Public Health Association (24), but using the EIL dissolved oxygen probe model no. 8012 to measure oxygen concentration (Electronic Instruments Ltd., Richmond, Surrey,UK.)
Samples of raw and treated slurry were centrifuged at 10,OOOg for 20 minutes at 10°C, as described by Hissett et al (25), to prepare the supernatants for the measurement of BOD. Determination. of. Total ..Organic A?ids, (TOA) TOA were determined, in the supernatant prepared as described by Hissett et al (25), by the method of Montgomery et al (26). The organic acids were esterified with acidified ethylene glycol. The esters were then reacted with hydroxylamine and the hydroxamic acids thus formed were converted to their ferric complexes and their concentrations were determined by optical density measurements at 500u. Gas-liauid Chromatography of VFA and TIP A Pye series 104 gas-liquid chromatograph (Pye Unicam Ltd.,Cambr idge,UK.) and a Packard model 439 gas-liquid chromatograph (Packard Instrument Ltd.,Reading,UK.) with a Hewlett Packard HP 3390A reporting integrator (Hewlett Packard Ltd.,Cheshire,UK.) were used. Determination of VFA were by direct injection of acidified supernatant samples into a 2.15 m long by 4mm ID packed column of 5% FFAP on Cnromosorb G.AW.DMCS 80/100 mesh. Determination of TIP were by direct injection of solvent extracts, of whole samples (30), into a 25 m long by 0.23 mm ID capillary column of WCOT fused Silica with liquid phase CP sil 5CB (United Technologies Packard,Reading,UK.). Odour Assessment The odour panel consisted of 21 College Staff who had previous odour panel experience (5) and 5 staff who were later recruited and whose performances at slurry odour assessments were judged to be similar to those of the original panellists. The method of assessment of odour offensiveness was based on that of Sobel (19). 20 ml samples of slurry were tranferred to 60 ml black glass bottles as described by Williams (5). These were handed to panellists in their own offices or laboratories where they were already accustomed to the background odour and were least hindered by interruption. The panellists were shown a copy of Table I and asked to assign the odour offensiveness of each sample to a value between 0 and 5. They were specifically requested not to consider the samples relative strength compared with the other samples.
Thirty seven samples of raw piggery slurry (TS 29 g/l, standard deviation 3 g/l; COD 37 g/l, standard deviation 3 g/l) were assayed over a period of 3 years. These had offensiveness values of between 2.3 and 3.5 with a mean value of 2.9(definitely offensive) and a standard deviation Odour offensiveness = 1.453 LogSupernatant BOD + 2.320 (2) Odour offensiveness = 2.378 LogTOA + 2.327 (3) Supernatant BOD and TOA concentrations are expressed as g/l. The mean variance of the offensiveness from the regression line and the standard error of the supernatant
and 0.50 respectively. Scatter diagrams of VFA (Figure 3) and TIP (Figure 4) concentrationsag ainst odour offensiveness of slurry are illustrated. The relationship of VFA concentration to odour offensiveness of slurry was only slightly improved by subtraction of the acetic acid concentration from the total VFA. Supernatant BOD^ and TOA The average supernatant BODc of raw piggery slurry was 2.349 g/1 (standard deviation 0.470) and that of TOA was 1.533 g/1 (standard deviation 0.213). The mean ratio of TOA to supernatant BOD^ was 0.666 (standard deviation 0.105). During continuous culture treatment supernatant BOD and TOA were oxidised.The ratio of TOA to supernatant of the
The relationship of VFA and TIP concentrations with odour offensiveness showed that at the lower offensiveness ratings the concentrations were too scattered to provide sufficient reliable indicators of odour offensiveness. This was also shown by Williams (5). REFERENCES 1 JONES, K.B.C. 1977 U.K. problems and Legislation relating to odour control. Agriculture and Environment ^ (2,3) 245-258. 2 PEAKIN, F.H. 1979 Odours in agricultural practice, prevention and control. Agricultural Research Council, London. 3 BAINES, S., McLARTY, R. & MILLS, P. 1980 Odour in agriculture. The Scottish Agricultural Colleges publication No 57, May 1980. BATTERSBY, S.A. Institute of Environmental Health Offices Agricultural Odour Survey. April 1980 - March 1981. MAFF Farm Odour Group paper FOG/14 (unpublished). 5WILLIAMS, A.G. 1981 The biological control of odours emanating from piggery slurry. PhD Thesis, University of Glasgow, 1981. WILLIAMS, A.G. 1984 Indicators of piggery slurry odour
BEDBOROUGH, D.R. & TROTT, P.E. 1979 The sensory measurement of odours by dynamic dilution. Report No LR 299 (AP). Warren Springs Laboratories, Stevenage. 13 BARTH, C.L., HILL, D.T. & POLKOWSKI, L.B. 1974 Correlating odour intensity index and odorous components in stored dairy manure. Transactions of the American Society of Agricultural Engineers 17(4), 742-4, 747. 14 SCHAEFER, J. 1977 Sampling, characterisation and analysis of malodours. Agriculture and Environment, 3(2,3), 121-128. 15 SCHAEFER, J. 1980 Development of instrumental methods for measuring odour levels in intensive livestock buildings. In: Effluents from Livestock (Gasser, J.K.R. (Ed.)). Applied Science Publishers, London pp 513-535. 16 KOWALEWSKY, H.H., SCHEV, R. & VETTER, H. 1980 Measurement of odour emissions and immissions. In: Effluents from Livestock (Gasser, J.K.R. (Ed)), Applied Science Publishers, London pp 609-626. 17 HARPER, R., BATESMITH, E.C. & LAND, D.G. 1968 Odour description and odour classification. J & A Churchill Ltd., London. 18 BELL, R.G. 1970 Fatty acid content as a measure of the odour potential of stored liquid poultry manure. Poultry Science, 49, 1126-9. 19 SOBEL, A.T. 1972 Olfactory measurement of animal manure odours. Transactions of the American Society of Agricultural Engineers. 15(4), 696-699 and 703. 20 AMOORE, J.E., VENSTROM, D. & DAVIS, A.R. 1968 Measurement of specific anosmia perceptual and motor skills 26, 143-164. 21 SPOELSTRA, S.F. (1980) Origin of objectionable odorous components in piggery wastes and the possibility of applying indicator components for studying odour development. Agriculture and Environments(3), 241-260. 22 OWENS, J.D., EVANS, M.R., THACKER, F.E., HISSETT, R. & BAINES S. 1973 Aerobic treatment of piggery waste. Water Research 7 1745-66. 23 EVANS, M.R., HISSETT, R., SMITH, M.P.W., THACKER, F.E. & WILLIAMS, A.G. (1980) Aerobic treatment of beef cattle and poultry waste compared with piggery waste. Agric. Wastes 2, 93-101.
APHA 1971 Standard methods for the examination of water and waste water 13th Edition. American Public Health Association Inc., New York. 25 HISSETT, R. , EVANS, M.R. & BAINES, S. 1975 The use of respirometric methods for assessing the biodegradability of different components of agricultural wastes. Progress in Water Technology 7(2) 13-21. 26 MONTGOMERY, H.A.C., DYMOCK, J.F. & THOM, N.S. 1962 The rapid colourimetric determination of organic acids and their salts in sewage sludge liquor. Analyst 87, 949-955. 27 BAINES, S. & EVANS, M.R. 1985 Aeration and odour control by heterotrophic and autotrophic aerobes. EEC/FAO, Silsoe, U.K. 15-19 April 1985. 28 SPOELSTRA, S.F. 1977 Simple phenols and indoles in anaerobically stored piggery wastes. Journal of the Science of Food and Agriculture 28^ 415-423. 29 SPOELSTRA, S.F. 1979 Volatile fatty acids in anaerobically stored piggery wastes. Netherlands Journal of Agricultural Science 21_, 60-66. 30 Her Majesty's Stationery Office 1981 Phenols in waters and effluents. 31 Personal Communication with A.G. Williams, Farm Buildings Division, N.I.A.E., Silsoe, Bedford, U.K.
DUST IN LIVESTOCK BUILDINGS AS A CARRIER OF ODOURS J. HARTUNG Institute for Animal Hygiene of the Hannover School of Veterinary Medicine, Blinteweg 17p, 3000 Hannover 71, FRG Summary The dust of animal houses originates mainly from the feed (80-90%), the bedding material, the manure (2-8%),and the animals (2-12%) themselves. It consists substanticaily of organic matter. The factors determining the amount of dust in confinements include animal activity, temperature,rel-ative humidity, ventilation rate, stocking density and volumetric air-space per animal, feeding method, and na ture of the feed. This dust originating from various sources can carry gases, vapours and odours. The analysis of dust-borne trace gases is usually done by solvent ex traction followed by gas chromatography. At least 60 com pounds belonging to different chemical groupings were i-dentified in the dust from animal houses. Volatile fatty acids and phenolic/indolic compounds were found to con tribute mostly to the strong, typical odour of animal houses. Main components in these groups are acetic acid and p-cresol, respectively. In the dust from pig houses qualitatively and quantitatively nearly the same volatile fatty acids and phenols are found than in the air or in the slurry. One m3 of the exhaust air from a 500 head pig fattening unit can contain dust-borne 6.27 pg volatile fatty acids and 2.76 yg phenolic/indolic compounds. The concentration of odours on the dust particles seems to be much greater than in an equal volume of air. Filtering the dust from the exhaust air can reduce the odour emission from animal houses up to 65%. Another way to reduce the dust-borne odour emission is to avoid the release of dust in the animal house by wet feeding, vacuum cleaning or showeri ng. 1. INTRODUCTION Dust in animal houses is an atmospheric contaminant of the environment of the animals (1). It is an important carrier of microorganisms (2), (3), (4), and can influence the perform ance and health of animal (5), (6), (7), and man (8),(9),(10). In addition the dust of animal houses was supposed to play an essential role in the transport of trace gas and odour inside and in spreading of odorous gases outside of the animal house (11), (12), (13), (14). This paper reports on the aspects of dust formation in livestock buildings, the material composition of the dust, the
emission of dust-borne odourants like volatile fatty acids (VFA) and simple phenols and indoles from piggeries, the impor tance of particle-borne odours, and the possibilities of con trolling dust-borne odours. 2. ORIGIN, NATURE AND RELEASE OF THE DUST It is estimated that the dust in animal houses originates mainly from the feed (15 ), (16 ), (17 ), the bedding material (18), (19), the manure (20) and the animals themselves (21),(22). Relevant values are rare. Table I shows that feed and bedding, when used, are the predominant sources of dust in pig and hen houses. Dust from animal houses consists mainly of organic matter (23). The preferred technique for investigating both the mate rial composition of the dust and feed stuff is the WEENDER An alysis Technique (24). Table II shows the composition of dust from pig and hen houses compared to the feed fed. The differ ences in the protein content between dust and feed support the opinion that an important part of the dust originates from feathers, hairs, and skin cells of the animals. The release of the dust is caused by the activity of ani mals or man or the function of technical equipments in the an imal house. Feeding, particularly dry feeding (25), as well as bedding and cleaning activities, the use of different systems of feed distribution, manure removal and ventilation (26) can increase the dust level in the air of animal houses consider ably (27). Figure 1 gives an example of the relation between the amount of dust in the air and different activities based on values as reported by CERMAK and ROSS (27) for poultry houses. In the course of a day the dust level in animal houses varies considerably. Mostly feeding increases the dust concen tration in the air as demonstrated in Figure 2 (22). However, within 30 to 120 min the "normal" background level is reached again (16),(22). The figure shows that even before the feed is distributed, the activity of the animals increases the dust concentration in the air considerably. Table III shows the influence of rel . humidity, pen vol ume, feeding system and air flow on the number of dust parti cles and weight of settled dust in an experimental piggery.The essential influence of animal activity on the formation of dust is shown by the fact that self-feeding results in significant ly greater atmospheric dust concentration (particles/volume of air) than does floor-feeding. However, a significantly greater amount of settled dust is associated with floor feeding. Prob ably, the self-fed pigs spend much more time eating than the floor-fed pigs. The intense activity of the pigs during floor feeding results in a great deal of visible dust for only a pe riod of time, while the self-fed pigs may play with the excess feed (28),(17). These studies indicate that the factors deter mining the amount of dust in confinements include animal ac tivity, temperature, relative humidity, ventilation rate,stock ing density and volumetric air-space per animal, feeding method, and nature of feed. This dust originating from various sources can carry gases, vapours and odours (7).
DUST-BORNE TRACE GASES AND ODORANTS The analysis of dust-borne trace gases requires their i-solation from the dust particles. Procedures for the isolation and characterization of trace gases and odorants in the dust from pig houses are given by SCHAEFER et al. (29), HAMMOND et al.(30) and TRAVIS and ELLIOTT (31). Alcoholic solvents were found to be effective for the extraction of volatile fatty ac ids and phenols from the dust of hen (32) and pig houses (33), (34). Today, gas chromatography is usually used for the sepa ration and identification of the trace gases. Table IV gives a literature review of compounds identified in the dust of pig houses. There are only very few reports on investigations on the dust from hen houses (32). Most of the odours coming from livestock production units are associated with the biological degradation of the animal wastes (35), the feed and the body odour of the animals (1). Volatile fatty acids and phenolic compounds were found to con tribute mostly to the strong, typical odour of animal houses by the help of sensory evaluations parallel to the chemical analysis (29),(30). Table V gives quantitative values of volatile fatty acids and phenolic/indolic compounds found in the aerosol phase and in settled dust of piggeries, respectively. The results from the aerosol phase coincide, particularly as far as acetic acid is concerned. For the investigations of the settled dust the coefficients of variation (CV) and the relative values (%) characterizing the percentage of the single compounds as part of the total amount are quoted. The values are corrected with the dry matter content of the dust. Main components are acetic acid and p-cresol, respectively. Table VI compares results from air, dust and slurry in vestigations on VFA and phenolic/indolic compounds in piggeries. Relative values are used. When comparing the results derived from investigations on dust, air or slurry it is necessary to use relative values because of the different dimensions, for experience shows that in spite of large quantitative differ ences between two samples within the group of carboxylic acids and within the group of phenolic/indolic compounds the propor tions of the components remain rather stable (36). In the group of VFA acetic acid is the main component in air, dust, and slurry followed by propionic and butyric acid. The other three acids amount to less than 25%. In the group of phenols/ indoles p-cresol is the main component in the four cited in vestigations. However, it seems that straw bedding can reduce the p-cresol content; in this case phenol is the main compo nent , i nstead (37 ). 4. EMISSION OF DUST-BORNE VFA AND PHENOLS/INDOLES FROM PIGGERIES The investigations of dust from piggeries show that both VFA and phenols/indoles are present in a considerable amount. However, compared to the air-borne emissions calculated on the base of the results of LOGTENBERG and STORK (38) less than the tenth part (1/10) of phenols/indoles and about the hundredth part (1/100) of VFA are emitted by the dust, only. Table VII compares the dust-borne and air-borne emissions of VFA and
phenols from piggeries. The total amounts are given in addition to the amounts of butyric acid and p-cresol which are both known as intensively smelling compounds. The recognition odour threshold values of these two components are included, as well. Under the assumption of a dust concentration of 10 mg/m3 (7) one cubic meter of air from a pig house contains 6.27 pg dust-borne VFA and 2.76 |j g dust-borne phenol ic/i ndol i c compounds; 62.7 mg VFA and 27.6 mg phenolic/indolic compounds are emitted from a 500 pig fattening unit per hour at a medium ventilation rate of 20m3/70kg pig-h. When comparing the dust-borne concentrations of butyric acid and p-cresol with the odour thresholds it seems that the concentrations are too small to be relevant for an odour nui sance. However, if the dust is removed from the gas phase of the air from animal houses the odour disappears ( 39),(40) ,(14). This supports the opinion of HAMMOND et al. (40) that the odor is concentrated on the dust particles. The authors conclude from their data that the concentration^ of the two odorants bu tyric acid and p-cresol is about 4 • 10 greater on an aerosol particle than it is in an equal volume of air. Thus, an aero sol particle deposited on the olfactory organ carries odour equivalent to a much greater volume of air (40). These consid erations indicate that dust from animal houses should be taken into account in connection with odour emission/immission meas urements not only by chemical analysis but by sensory evalua tions using olfactometers without dustfilters, as well. 5. CONTROL OF DUST-BORNE ODOURS There are basically two ways of controlling dust-borne odours. An effective way seems to be the filtration of the air to remove the dust (41). VAN GEELEN (14) reports on the reduc tion of the odour emission from a broiler house with 15.000 animals of 65% by means of filter bags when filtering the ex haust air. However, the investments and running costs amounted to about DM 4.00 per 100 birds per year. The second way is to avoid the dust release in the animal house as far as possible. The following possibilities are recommended: - feeding intervals, no self-feeding (17) - pellets instead of meal feed - wet feeding instead of dry feeding (25) - vacuum cleaning, fogging and showering (22) A reduction of the dust content in the air of animal con finements bare not only the chance to diminish the odour emis sion from the animal houses but can have a positive influence on the animals' health and performance, as well. Acknowledgement The author wants to express his thanks to Dr. W.Heidmann, Chem-mical Institute for his help in preparing Table IV, Dr.G.Klink-mann for revising the English text, Mr. K.H. Linkert for doing the drawing, and Mrs. U. Arzt for typing the manuscript.
REFERENCES HONEY, L.F. and J.B. McQUITTY (1976). Dust in the animal environment. Research Bulletin 76-2, 1-66. (2 ACKEMANN, H.H. (1980). Quantitative Untersuchungen liber den bakteri-ellen Keimgehalt des Absetzstaubes in zwei Schweinemaststallen. Dtsch Tierarztl. Wschr. 87, 335-338. (3 LANGE, A., G. MEHLMRN, W. METHLING and V. NEUPARTH (1983). Dynamik der bakteriellen Kontamination des Staubes in Abferkelstallen. In: 5. Int. Leipziger Tierhyg. Symp., Leipzig, Sammelbd. d. Vortr. S. 137-142. (4 HILLIGER, H.G. (1984). Zur Bilanzierung der Bakterienflora in der Stalluft. Zbl. Vet. Med. B,31, 493-504. (5 MARTIN, H. and R.A. WILLOUGHBY (1972). Organic dust, sulfur dioxide, and the respiratory tract of swine. Arch. Environ. Health 25, 158-165. — (6 OWEN, J.E. (1982a). Dust - the problem and possibilities. Farm Bldg. Progress 67, 3-6. (7 CURTIS, ETC. (1983). Environmental management in animal agriculture. Iowa State University Press, Ames, Iowa. (8 PEPYS, I., P.A. IEMKINS, G.M. FESTENSTEIN, P.H. GREGORY, M.E. LASEY and F.A. SKINNER (1963). Farmer's lung: Thermophilic actinomycetes as a source of "farmer's lung hay" antigen. Lancet, 607-611. (9 BUTIKOFFER, E. and A.L. de WECK (1969).Huhnerzuchterlunge. Dtsch. med. Wochenschr. 94, 2627-2631. KOSTERS, J. (198177 Stallstaub kann gefahrlich werden. DGS 33, 292-293. DAY, D.L., W.L. HENSEN and S. ANDERSON (1965). Gases and odors in confinement buildings. Trans. ASAE 8, 118-121. (12 BURNETT, W.E. (1969). Odor transport by particulate matter in high density poultry houses. Poultry Sci. 48, 182-185. (13 WEURMAN, C. (1975). Vergleich zweier Methoden fur die Messung von Ge-riichen. VDI-Bericht 226, 135-139. VDI-Verlag GmbH Dlisseldorf. (14 VAN GEELEN, M. (1983). Stankproblemen bij siachtkuikenhok zijn even-tueel op te lossen. Pluimveehouderij 13, 12-13. (15 CURTIS, S.E., J.G. DRUMMOND, D.J. GRUNLOH, P.B. LYNCH and A.H. JENSEN (1975). Relative and quantitative aspects of aerial bacteria and dust in swine houses. J. Animal Sci. 41, 1512-1520. (16 BRESK, B. and J. STOLPE (1975). "Der EinfluB des Staubes in industrie-maBigen Schweineproduktionsanlagen auf die Lei stung und Gesundheit der Tiere. Monatsh. Veterinarmed. 30, 572-576. (17 HONEY, L.F. and J.B. McQUITTY (1979T. Some physical factors affect ing dust concentrations in a pig facility. Can. Agric. Engineering 21, 9-13. (18 MHO , C.A. et al. (1969). Dust production of poultry litter materi als. Auburn Univ. Agr. Exp. Sta. Circ. 169. (19 MATTHES, H. (1979). Art und Zusammensetzung der Luftverunreinigungen in der Nutztierhaltung und ihre Wirkung in der Stallumgebung. Dtsch. Tierarztl. Wochenschr. 86, 262-265. MAN, C.,L. CERNEA and T. BUHATEL (1971). Examenal calitativ pulberi-lor din aerul adapsturilor pentru pasari. Lucrari stiintifice, seria medicina veterinara 27^ , 321-329.
KOON, J., J.R. HOWES, W. GRUB and C.A. ROLLO (1963). Poultry dust: Origin and composition. Agr. Eng. 44, 608-609.
(22) NILSSON, C. (1982). Dust investigations in pig houses. Swedish Uni versity of Agricultural Sciences, Department of Farm Buildings, Di-vison of Farm Building Constructions, Lund. Rapport 25, pp 93. (23) HILLIGER, H.G. (1966). Gravimetrische Staubmessungen in Stallen. Zbl. Vet. Med. B, 13, 698-708. (24) PALOHEIMO, L.T1969). Weender Analyse. In: W. LEMKEIT, K. BREIREM and E. GRASSMANN (Hrsg.). Handbuch der Tierernahrung, Bd. 1, S.164-171, Verlag Parey, Berlin, Hamburg. (25) HELEN, M. (1984). Einige Ursachen fUr die Variationen der Staubkon-zentration im Mastschweinestal1. In: Symposium der Internationalen Gesellschaft fUr Tierhygiene, Hrsg.: Deutsche Veterinarmedizinische GeselIschaft, 28-30. (26) NAKAUE, H.S., J.K. KOELLIKER, D.R. BUHLER and G.H. ARSCOTT (1981). Distribution of inorganic elements in poultry house dust. Poultry Sci. 60, 1386-1391. (27) CERMAK, J.P. and P.A. ROSS (1978). Airborne dust concentrations as sociated with animal housing tasks. Farm Buildg. Progr. 5J_, 11-15. (28) BUNDY, D.S. and E.T. HAZEN (1975). Dust levels in swine confinement systems associated with different feeding methods. Trans. Amer. Soc. Agric. Eng. J8, 137-139. (29) SCHAEFER, J., J.M.H. BEMELMANS and M.C. Ten NOEVER DE BRAUW (1974). Onderzoek naar de voor de stank van varkensmesterijen verantwoordi-lijke componenten. Landbouwkund. Tijdschr., pt 86-9, 228-232. (30) HAMMOND, E.G., C. FEDLER and G. JUNK (1979). Identification of dust-borne odors in swine confinement facilities. Trans. ASAE 22, No. 5, 1186-1189 & 1192. (31) TRAVIS, T.A. and L.F. ELLIOTT (1977). Quantitation of indole and scatole in a housed swine unit. J. Environ. Qual. 6 (34) HARTUNG, J. (1985). Gas chromatographic investigations of swine house dust on odorous compounds. Environmental Technology Letters 6, 21-30. (35) SPOELSTRA, S.F. (1978). Microbial aspects of the formation of malo dorous compounds in anaerobically stored piggery wastes. Wageningen, Landbouwhogeschool, Diss., pp. 91. (36) SCHAEFER, J. (1977). Sampling, characterization and analysis of mal-odours. Agric. Environm. 3, 121-127. (37) HARTUNG, J. and E. R0KICK.T (1984). Zum Vorkommen phenol art iger Ver-bindungen im Staub von Schweine- und HUhnerstall. Zbl. Bakt. Hyg., I. Abt. Orig. B, J_79, 431-439. (38) LOGTENBERG, M.Th. and B. STORK (1976). Het ontwikkelen van meetme-thoden voor het bepalen van de stank van ventilatielucht van mest-varkensstallen.Rapport de Centraal Technisch Instituut TNO, Zeist/ Holland. Ref.no: 76-06 054, Dossier: 01-4-40130. (39) WILLSON, G.B. (1971). Control of odours from poultry houses. ASAE Symp. Livestock Waste Management, Columbus/Ohio, 19.-22.4.1971. (40) HAMMOND, E.G., C. FEDLER and R.J. SMITH (1981). Analysis of particle -borne swine house odours. Agric. and Environment 6, 395-399. (41) OWEN, J.E. (1982b). Dust - Filtration solutions an? their cost. Farm Building Progress 68, 19-23.
407-410. (32) HARTUNG, J. (1983). Spurengase im HUhnerstallstaub. In: 15. Kongress d. Dtsch. Veterinarmed. Ges., Bad Nauheim, Ber. S.246-250 (Fortschr. Veterinarmed. 37). (33) AENGST, C. (19M). Zur Zusammensetzung des Staubes in einem Schweine- maststall. Hannover, Tierarztl. Hochsch., Diss.
(42) CORMACK, D., T.A. DORLING and B.W.J. LYNCH (1974). Comparison of techniques for organoleptic odour intensity assessment. Chem. Ind. (London) no.2 , 857-861. (43) LEONARDOS, G., D. KENDALL and N. BARNARD (1969). Odor threshold de termination of 53 odorant chemicals. J. Air Poll. Contr. Assoc. 19, 91-95. dust m g/rrr
[Al I l.ul L .,1—L
Figure 2: Dust concentration (mg/m3 ), ventilation rate (m3/ pig h), temperature (°C), and relative humidity dust p ig 1) he n 2) origi n o tin g without bedding battery bedding from •/. % •/. feed 8 0 -9 080 -9 0n.r. bedding --5 5 -6 8an im a ls 5 -1 1 2 -1 2 m an u re n.r. OOiao1i
during 24 h in a pig fattening house from NILSSON (22). The arrows (?) indicate time of feeding. Table I: Predominant sources of the dust in pig and hen houses from 1)= HONEY and McQUITTY (17) and 2)= MAN et al. (20). n.r. = not reported.
Table II: Material composition of dust and feedstuff from pig and hen houses. Results from 1)= AENGST (33), 2) = HARTUNG (34), 3)= KOON et al. (21), 4)= HAR-TUNG (32). n.r. = not reported. pig hoi js e (no 1 bedding] hen hou se(b components dust” dust2) feed1) dust3> dust4) •/. •/. •/. •/. dry matter 87 87 88 92 89 92 crude protein 24 24 19 60 50 17 fa t4549 10 15 crude fibre 3554 nr. 2 ash 15 nr 5nr nr. 4 Table III: Number of particles collected by the Andersen Sampler and weight of settled dust in an experi mental piggery at different conditions from HONEY and McQUITTY (17) number of particles / 0.028 n3 treatment particle size settled dust 7 -16 jjm <5jjm g/rrfd ay rel humidity (low ) 51 370 119920 13.42 rel humidity (high) 32450 85230 10.03 pen volume 22.1 nf 38200 86420 12.38 pen volume 11.1m3 45630 118 730 11.08 floor feeding 42 770 883 90 15.85 self feeding 41050 116760 7.62 a irflo w 595rr?/h 38650 93820 12.08 a irflo w 297rr?/h 45170 111 330 11.37 average 41900 102580 11.37 ^ TJa
Table IV: Volatile compounds identified in the dust of swine confinement units Hydrocarbons Ketones Hexane 1) Acetone 4) «t-Pinen 1) Butanone 4) Limonen 1) Pentanone 4) 3,7,7-Trimethyl -Octanone 4) bicyclo (3,1,1)-l-0ctene-3-one 4) 2-Hepten 1) Benzene 1) Acids Toluene 1) Acetic 4)5)6) Alcohols Propionic 4)5)6) i-Butyric 1)5)6) 1-Pentanol 1) Butyric 4)5)6) 1-Heptanol 1) 1-Valeric 5)6) 4-Methylcyclo-Valeric 4)5)6) hexanol 1) Hexanoic 4)5) 2-Ethylhexanol 1) Heptanoic 4) Octanoic 4) Phenols Nonanoic 4) Decanoic 4) Phenol 1)3)6) Undecanoic 4) p-Cresol 3)4)6) Dodecanoic 4) o-Cresol 1) Laurie 3) p-Ethylphenol 3)6) Tridecanoic 4) o-Ethylphenol 4) Tetradecanoic 4) m-Ethylphenol 4) Benzoic 4) Phenyl acetic 3)4) Indoles 3-Phenyl propionic 3) Hydrocinnamic 4) Indole 2)6) Skatole 2)3)6) Miscellaneous Compounds Aldehydes 2-Pentylfuran 3) Vanillin 3) Butanal 4) 2-Butenal 4) Pentanal 4) 2-Pentenal 4) Hexanal 1)3)4) 1) = WEURMAN (13) 2-Hexenal 4) 2) = TRAVIS and ELLIOTT (31) Heptanal 1)3) 2-Heptenal 4) 3) = HAMMOND et al. (30) 2.4-Heptadienal 3)4) 4) = HAMMOND et al. (40) Nonanal 3) 2-Nonenal 3) 5) = AENGST (33) 2.4-Nonadienal 3)4) 6) = HARTUNG (34) Decanal 4) 2.4-Decadienal 3)4) Benzaldehyde 1)4)
Table V: Amounts of odourous constituents in the aerosol phase (M 9 / 9) arid in the settled dust (^ig/g) of swine house air from 1)= HAMMOND et al. (40), 2) = AENGST (33), and 3) = HARTUNG (34). - indicates not reported. quantitative results quantitative results in n g /l aerosol in jug /g settled dust compounds 1) 2 ) 3) CV relative 0.3 62 4 84 9.9 0.2 --- 11 488 40 skatole -- 43.8 54
relative values acetic acid propionic acid iso-butyric acid butyric acid
Table VII: Amounts of volatile fatty acids and phenolic/indolic compounds emitted from piggeries by dust and air, respectively. The recognition odour thresholds for butyric acid and p-cresol are included. 1)= HARTUNG (34), 2) = LOGTENBERG and STORK (38), 3)= CORMACK et al. (42), 4)= LEONARDOS et al. (43). components em itted odour dust-borne air-borne threshold P91"? } 2J
ijg/m 3 pg/rr? volatile fatty acids butyric acids phenolics / indolics total