Challenging odour regulations
Phil Longhurst looks at the difficulties of controlling odours in wastewater treatmentSmell problems have a long lineage with wastewater treatment designers. London's great stink prompted the most significant development in the capital's sewerage network. Odour from the highly polluted River Thames was so offensive that by the 1850s the Houses of Parliament could no longer function properly.
Curtains soaked in lime were hung at the windows facing the river and relief was only assured after 1864 with the completion of London's main sewerage system. Sir Joseph Bazalgette's contribution, as chief engineer at the time, was significant in achieving health benefits, disease control and improvements in river ecology. The clean-up, however, had been driven by public amenity and the result was that people could again use their environment. Understandably, odour control is now secondary to the majority of process performance requirements. Odour is not identified for capital funding as a performance requirement by the regulator and can reasonably be expected to be a low priority on the list for process performance specifications. Yet, after 240 years, the control of odours is returning as a principle consideration for the design and operation of WwTWs.This shift is not driven by a search for competitive advantage but in direct response to complaints and, in some cases, financial survival. Preventing odours from wastewater treatment relies on ensuring one of these three stages can not occur:
- malodorous compounds being present or forming in the liquid phase,
- odorous compounds being released in to the gas phase,
- receptors being exposed to odorous compounds in the gas phase.
When the flow of effluent is stable and its composition unchanging, this seems straight forward. For most treatment plants this is the exception to the rule. In recent years, odour problems have resulted in notable diversions for operators, ranging from bad publicity to prosecutions.
For the most part, the journey has tied up significant time for senior staff dealing with the problem. Large-scale investment at short notice has often been required to correct problems with water company staff facing confrontations with angry locals, action group meetings or the press. Companies experiencing these pressures can end up working to higher standards if fitting installations retrospectively than if considered in the first place. Similarly, they may face requirements or are cautious to install stringent odour control requirements at future treatment works. Currently no laws specifically regulate odours but because odours are a nuisance pollutant they are dealt with by nuisance legislation, principally in the UK, the Environmental Protection Act 1990. Every local authority has a duty under Section 79 to detect any statutory nuisances that ought to be dealt with under Section 80. Here the authority serves an abatement notice requiring the abatement of the nuisance or prohibiting or restricting its occurrence.
This may require the execution of works as well as other steps necessary. A person served with such a notice can appeal to a magistrates' court within 21 days to avoid the maximum fine of £20,000 for non-compliance. A defence is to prove the best practical means were used to prevent, or to counter the effects of, the nuisance. Private prosecutions may also be brought on the grounds of a statutory nuisance. Defra's consultation document, issued in December 2002, put forward four possible options for considering alternative approaches for the control of odour and nuisance from WwTWs:
- no additional legislation, a code of practice for the water industry,
- amend EPA 1990 to extend statutory nuisance provisions to WwTWs,
- extend LAPPC provisions under to apply to WwTWs,
- extend IPPC provisions to include WwTWs.
But while an outcome is pending, case law and guidance continues to be used to determine acceptable limits to exposure. So why are odour limits for process design and regulation difficult to define? Certainly, there are no easy and cheap measures.
Hydrogen sulphide is a common and measurable component at source. Downwind measures can also be detected, but this does not represent all forms of odour on treatment plants. Sensory panel measurements of odour concentrations (Sneath, 2001) are very effective but costly. They are not in real-time and must be taken from the source to prevent contamination from background odours. Certainly, measurement is problematic but it is improving in reliability and range (van Harreveld et al, 1999). So why has control and regulation not become more straight forward? Two challenges exist - human response to the perception of odour and ensuring an adequate level of prevention and response.Odour perception
Most odour discharged to the atmosphere consists of a complex mixture of components. Human sensory responses to the individual components of such mixtures vary considerably from compound to compound, and from person to person. For all the human senses, including smell, a general approach to assessing the relationship between the magnitude of a sensation and the intensity of stimulus can be used. A method referred to as category estimation can be derived from Fechner's Law (Stevens, 1960) where it is proposed equal ratios of stimulus lead to equal differences between perceived intensity.
This phenomenon is recognised for other sensory stimuli such as noise where the decibel (log) scale is used. In the same way the sense of smell is a log response curve. Thus perceived intensity (I) is, in part, a linear function of the logarithm of odour concentration (C):I = k1 (log10C) + k2
(where k1 and k2 are constants for a particular group of panellists within an experiment). Perceived odour intensity can decrease rapidly for continuous exposure. This is referred to as adaptation. Sensitivity to odour is also recovered when the exposure is removed. Both these processes, adaptation and recovery, operate over short timescales and are not yet well understood. It follows if attempting to achieve a reduction in impact from odours then the scale of recognition will be determined by the sensory scale, a log scale.
That is, log scale improvement is needed to ensure recognition of a reduction. So are some odours worse than others? In short, the key is pleasant and unpleasant. Both et al (2003) recently showed where an unpleasant odour could be recognised, that once annoyance was established, reducing high concentrations to a point where the odour could still be detected was unlikely to remove the problem. This study took an extensive series of field measurements and interviews using a standardised questionnaire across six installations, each with differing levels of unpleasantness - hedonic tones. Panels were used to assess the odour impact alongside annoyance to residents. A new method to measure odour intensity and hedonic tone in the field was developed to produce reliable and reproducible results. The odour frequency, recorded as a system of 'odour hours', was used to predict odour annoyance caused by unpleasant or neutral odours.
This in turn was matched to the reported exposure. Results showed in the case of pleasant odours hedonic tone has an abundantly clear effect on the dose-response-relationship between odour frequency and annoyance. Pleasant odours have a significant lower annoyance potential than unpleasant/neutral odours. Intensity, for example, the perceived strength of the odour was shown to have no additional influence on this relationship. If odours are recognisable they can cause annoyance. Preventing or evaluating the impact of odour on a community needs three forms of information - accurate emissions data, dispersion modelling and an odour annoyance index derived from the temporal variation of odour concentrations. Any good 'source, transfer and receptor' model should provide us with knowledge of how much is too much at any point in the pollution pathway. Currently, the lack of knowledge about odour emissions, especially from time-varying sources, limits our ability to make accurate predictions from common malodorous processes.
While this is a focus of current research there is an ongoing need to understand the variation of emissions from processes. Techniques for dispersion modelling have progressed but their data dependence often results in proxy data being used for emissions in place of direct sampling.
This leaves us seeking information on performance from the site itself. Commonly, operators are aware of processes that are smelly but this information is rarely recorded, analysed or applied. Benefit can be gained from auditing the treatment process to gain the data to construct risk grids. This involves a combination of site visit-based assessments of potential odour problems and the use of stakeholder data. Auditing relies on collecting data on the odour nuisance potential of a site. Qualitative data from site staff and complaints is collated where, in many situations, this data is generated in the form of complaints or staff observations. Observations are encouraged from site visitors and on some sites, 'odour champions' have been appointed to raise awareness of both the importance of odour for the company and the central knowledge points for the stakeholder and other data. Data obtained from walk-over surveys can often identify a number of potential odour problems and offer easy fixes. Examples such as top-loading of sludge tanks or poor desludging of primary sedimentation tanks occur alongside housekeeping measures. Qualitative and quantitative data on odour treatment assets, performance, costs, the design basis and required maintenance are also relevant.raising awareness
Centralised storage and access to this information can then inform decisions on appropriate technology for future abatement schemes. By raising awareness of odour issues two immediate benefits arise - firstly, information is accessible to senior staff, meaning they are aware of the importance of odour issues and potential problem sites; and secondly, the information is retained when staff move sites or leave. Hence informed investment decisions can be made.
Figure 1 illustrates how the information gained from understanding the odour impact of a process could be assessed. By investing in improvements in process 'A' it is anticipated impacts can be reduced from significant incidents occurring on an almost weekly basis to lesser events occurring on an annual or seasonal basis. While qualitative in nature, this allows comparative judgements to be made about investments across a site or sites. The consequences of this can be summarised to point to the demands that future odour regulations will have to meet:
- a difference exists between the limits that are appropriate for process design limits and lower concentrations showing that odours can be detected,
- process designers need guidance to work to specific emission limits. These need to take account of the concentration of odours that will not establish annoyance from unpleasant odours but may be detectable,
- where annoyance has been established, significant reductions in process emissions are needed to overcome this. Achieving low-level concentrations needs to be determined by design. Knowledge of the odour source is essential and best available techniques should be the guide to prevent these. Regulations can play a role in establishing how this assessment and design guidance should take place,
- where elements of the population are sensitised to emissions, it may not be possible to ensure no further exposure to offensive odours. Where this occurs, communicating the level of reduction that will reliably be achieved and ensuring this reduction is detectable is essential.
In short, our own impact on sensory perception is responsible for creating a moving threshold of sensitivityReferences: Both, R Sucker K, Winneke G, Koch E (2003). Odour intensity and hedonic tone - important parameters to describe odour annoyance of residents? Proceedings from the 2nd IWA International Conference on Odour and VOCs, Singapore, September 14-17. Sneath, RW (2001). Olfactometry and the CEN standard prEN 17325. Odours in wastewater treatment - measurement, modelling and control. Stuetz R, Frechen F-B (Eds), IWA Publishing, London, p.95-119. Stevens, SS (1960). The psychophysics of sensory function, American Scientist, volume 48, 226-253. van Harreveld A, Heeres P, Harssema H. (1999). A review of 20 years of standardization of odour concentration measurement by dynamic olfactometry in Europe. Journal of the Air and Waste Management Association 49(6) 705-715.