| Forside | | Indhold | | Forrige | | Næste |

Risikovurdering af anvendelse af lokalt opsamlet fæces i private havebrug

Summary and conclusions

Introduction

Closing the nutrient cycle in urban areas is a major task for the future. Faeces and urine, as well as mixed sewage products need in this context to be seen as resources rather than as wastes. Hazards associated with recycling of these products include pathogens and pharmaceuticals as well as other organic contaminants/compounds and heavy metals. By treatment it is possible to reduce the number of pathogens, and pharmaceuticals and organic compounds may be degraded to some extent. Heavy metals are, however, not possible to eliminate.

From a hygienic risk perspective faeces should always be considered to contain pathogens. There are a wide range of pathogens, mainly causing gastro-intestinal infections, that may be excreted. When recycling other fractions such as urine or greywater, the potential faecal contamination and related pathogens may constitute the main risk. To evaluate and compare different sanitation systems including reuse of the waste products microbial risk assessment is a valuable tool that allows prospective studies. Epidemiological investigations often require larger study populations, and are mainly applicable when the practises are already in place.

Pharmaceuticals are, by nature, biologically active compounds where some are used and emitted in large quantities from society e.g. antibiotics and compounds for treatment of life-style related diseases. Pharmaceuticals are excreted with faeces and urine after normal therapeutical use and if the faeces or urine is re-cycled, pharmaceutical residues will end up in the environment. It is often anticipated that pharmaceuticals are degradable by microorganisms, since they are transformed by metabolisms in man, but according to the available data biodegradation is actually limited for some groups of compounds. Daily use of a certain type of drug may even lead to very high concentration of the compound in faeces (mg to g/kg faeces). The environmental hazard of pharmaceuticals and exposure assessment of pharmaceuticals to the environment have previously been addressed.

In addition to pharmaceuticals, humans are exposed to numerous other chemical substances, e.g. heavy metals and dioxins, which are present at low levels in food, water and ambient air. These persistent micro-pollutants of high environmental concern are also excreted with faeces but the hazard from the local handling of this material has not previously been assessed.

The case scenarios

The current project was aimed at assessing the risks both to humans and the environment from pathogens and pharmaceutical residues associated with usage of faeces within private households. The faeces were collected from dry urine-diverting toilets in single family households and used in their own gardens as a fertiliser. The faeces were only treated by means of storage prior to the application in the garden and, thus, the material was neither fully stabilized, nor were the pathogens fully inactivated. The following scenarios were evaluated:

1. Application of the material after storage for 0 months

2. Application of the material after storage for 6 months

3. Application of the material after storage for 12 months

4. Application and incorporation of the material after storage for 6 months

5. Application and incorporation of the material after storage for 12 months

In this case application means that the faeces is distributed evenly on top of the soil (i.e. mixed with the upper 1 cm of soil), while incorporation means that the material is actually worked into the upper layer of the soil, resulting in a faeces to soil ratio of around 1:100 compared to a ratio of 1:10 when the faeces is just applied on top of the soil.

Applied risk assessment procedures

Pathogens

Risk identification. For analysing the risks from pathogens, representatives from the various groups of microorganisms were chosen based on the following criteria:

• The organisms should be relatively common in Denmark
• The most persistent microorganisms should be included
• Organisms having low infectious doses should be included
• Some organisms causing infections potentially giving severe symptoms should be included
• There should be reasonable knowledge of the organism

As representatives for the bacterial group Salmonella and EHEC were chosen, and among the viruses, rotavirus and hepatitis A. To model parasites the protozoa Giardia and Cryptosporidium as well as the helminth Ascaris were chosen. All the pathogens included are transmitted through the faecal-oral route. Apart from hepatitis A they all cause gastro-intestinal infections, but further severe symptoms involving other organs may also occur.

Assessment of exposure. Each organism was modelled by means of distributions, i.e. probability density functions (PDFs), for incidence in the population, excretion and duration of infection as well as die-off in the storage container and die-off in the soil after application of the material in the garden.

Incidence data was mainly collected from Danish statistics, where Salmonella, EHEC, hepatitis A and Giardia are officially reported. Surveillance systems are however considered to underestimate the actual number of cases occurring in the population and the reported numbers were therefore corrected based on expert guesses, assuming higher underestimations for infections with less severe symptoms. Estimates of rotavirus and Ascaris infections in the population was made from extrapolation of the number of positive samples registered at Statens Serum Institut, Denmark whereas estimates of Cryptosporidium was based on incidence estimated by Hald and Andersen. The number of excreted pathogens i.e. their concentration in faeces, and the duration of excretion were collected from the literature. A critical step affecting the risks is the inactivation of the various pathogens in the collection and storage container (the container in which the faeces is collected is stored without addition of new material). Since available studies on the survival of pathogens in human faeces are limited, results from studies involving other materials such as animal manure and sewage sludge were also evaluated in order to establish PDFs for the inactivation. The collection and storage was assumed to take place indoors at a temperature around 20°C. The survival of microorganisms in soil is dependent on local conditions, e.g. soil type, moisture, UV-light and the naturally occurring microflora. To describe the inactivation in soil at an average ambient temperature of approx. 11°C, results from studies conducted in various types of soils were consolidated to create the PDFs. The PDFs representing the input data are summarised in table 1.

By combining the data the concentration of pathogens in the faeces after the different storage periods and in the soil after application and incorporation, including continuous inactivation, were calculated.

The actual human exposure was assumed to take place during one of three events, where accidental ingestion of small amounts of faeces or faeces and soil mixture may occur:

• Emptying of the container and distribution of the material
• Recreational activities in the garden
• Gardening

The faeces-soil intake was based on a literature study. Children are estimated to ingest around 200 mg of soil per day on an average with an absolute maximum of 5-10 g per day, occurring once every ten years by exposure each day. It was further assumed that adults ingest 15-50% of this amount , with a maximum of 100 mg per day. The container is emptied once a year, assuming that only adults are exposed. Modelling of the number of exposures through recreation resulted in a median value (50-percentile) of 3.5 times per week (during June-August), whereas 50% of the persons were exposed through gardening once a week (during May-September). An exposed "standard member" of the family was assumed to correspond to 25% child and 75% adult or older child (>6 years). The faeces to soil ratios mentioned above, being approximately equal to the 50-percentile, was used to create distributions for the composition of the faeces and soil mixture. It was further assumed that one exposure corresponded to two hours of gardening, occurring maximum two times per day.

Table 1. Input data used in the QMRA modelling. N (m, s) denotes a normally distributed PDF with mean m and standard deviation s. Some of the input variables are highly skewed to the right, meaning that there are higher probabilities of high numbers than of low numbers. This property of variable Y may be modelled by means of a log-normal distribution, i.e. ln(Y) is a normally distributed PDF.

T₉₀ = time for a 90% reduction

The dose-response relationships. For some pathogens dose-response relationships are relatively well defined and parameters and models can be found in the literature. Information is virtually missing for the susceptible parts of the population, such as children, the elderly or immuno-compromised and are not accounted for in the models. On the contrary, parts of the population are less susceptible, such as previously infected persons. In order to take this inherent variation into consideration the uncertainty of the parameters in the dose-response relationships must be included. As these studies are based on a small sample of individuals the uncertainties of the parameters were enlarged, thus attempting to model the overall variation in the exposed population.

Microbial risk calculation. The risk of infection in the QMRA was calculated applying 5000 iterations in the Monte Carlo simulations. Since many variables are described by means of distributions, calculations of both the realistic impact and also worst-case scenarios were possible, here translated to the 50-percentile and 95-percentile of the resulting distributions, respectively. Calculations were made for two main scenarios i) applying the incidence in the population (unconditional) and ii) assuming that one member of the family actually had an infection during the period of collection (conditional). Results are presented as probability of infection (P_inf) per exposure or per year and compared to an acceptable risk level of 10^-4.

Pharmaceuticals and micro-pollutants

Risk identification. Medicine metabolites were evaluated based on the criteria of frequent use in Denmark. The 25 most commonly used pharmaceuticals in the primary health sector in Denmark were assessed. For compounds where data was available a primary environmental hazard assessment for the soil environment and a health hazard assessment of ingesting soil that are polluted with pharmaceuticals were performed. In addition seven heavy metals (Cd, Cr, Cu, Hg, Ni, Pb, Zn) and dioxins (PCDD) were evaluated as representatives for toxic and persistent substances, considered as the micro-pollutants of highest concern in relation to application of faeces in gardens.

Exposure assessment. In the calculations it was assumed that a certain fraction of the drug was excreted with faeces as unchanged compound. Furthermore it was anticipated that compounds were persistent during the storage of faeces. For the exposure assessment of other micro-pollutants it was assumed that no degradation took place neither during storage nor in the soil after application.

Hazard assessment procedure applied in the calculations of risks. The environmental risk assessment was based on the EU draft guideline document for pharmaceuticals, and a risk quotient was calculated between the predicted environmental concentration (based on the scenarios described above) and the predicted no-effect concentration (PNEC) using worst-case assumptions. A default safety factor of 1000 was used to derive the PNECs. The additional safety factors for reduced data sets was not used. If the PEC/PNEC ratio exceed the ratio of 1 it implies that the compound have impact on the environment.

The human health hazard was assessed by an initial calculation of the amount of soil that should be ingested to be exposed for a Daily Defined Dose of the particular drug (DDD). Furthermore a number of cases where a "zero tolerance" perspective should be used were outlined. For the heavy metals and dioxins the environmental hazard assessment was simply based on comparison with existing quality standards for sludge and soil, and the human health assessment upon comparison with TDI-values (Tolerable Daily Intake).

Results and discussion

Calculation of microbial risks. The probability of infections (P_inf) are summarized in tables 2 and 3. The variations in the risk for infection depend on organism and were up to twelve orders of magnitude in a specific scenario. The risk from EHEC can be eliminated if the faeces are stored for one year. Many types of Salmonella are able to regrow in stored, but unstabilized materials, especially if partial wetting occurs. If the risk of regrowth of Salmonella can be ignored the risk of infection from Salmonella may also be ignored. Viruses and parasites generally have longer survival in the environment as well as lower infectious doses which resulted in high risks for rotavirus, the protozoa and Ascaris. It should however be noted that the calculations are based on reported die-off data. An extrapolation is often necessary to cover longer time periods than the actual measured period. Overestimations may occur if the survival curve does not follow first order die-off kinetics in the end of the estimated period. Ascaris, however, have a well documented prolonged survival time. According to table 2 the difference in risk between the conditional and unconditional scenario was 1-4 order of magnitudes and the difference between typical (50%) and worst case (95%) varied from none to 5 orders of magnitude depending on organism.

For viruses and parasites the risk of infection from fresh material is measured in percentages of the infected containers. The risk of infection is thus mainly dependent on the incidences of infections of the pathogen in question. Only after 12 months of storage (scenario 3 and 5) taking incidence into consideration the risks are below 10^-4 for all organisms, excluding Ascaris (P_inf 9 x 10^-4), when emptying the container and applying the material.

Table 2. Risk for infection when emptying the storage container (is done once per year and household) for scenario 1 (giving the highest risk) and for scenarios 3 and 5 (giving the lowest risk)

nr = negligible risk (<10^-14; calculation accuracy of MS Excel)

The typical (50%) yearly risks for gardening and recreation are quite high with a maximum of 4 x 10^-2 (0.04) for rotavirus in scenario 1 (0 months of storage; table 3). After storage for 12 months Rotavirus and Ascaris have the highest risk of infection. The risk of being infected by Ascaris is very high if the container was infected. As such it may be concluded that even after 12 months of storage the material must be incorporated properly in the soil in order to make the risks acceptable. These risks were calculated for exposure during May and June, respectively, thus being the highest during the period of activity.

Table 3. Yearly risk for infection (50-percentile representing the typical risk) during gardening and recreational activities in the garden. The numbers indicate the different scenarios for storage and usage of the faeces. Results obtained for the unconditional scenario, i.e. taking incidence into account

The facilities evaluated were based on the idea that the faeces would be composted. This process was however not taking place as indicated by the low temperatures recorded (0-20°C). Thus, the inactivation of pathogens will mainly occur due to time and ambient conditions such as moisture and by competition from other microorganisms. According to the literature the survival of pathogens, especially of protozoa and helminth eggs but also of viruses and some bacteria, can be long during these circumstances. This resulted in high risks for infection, with Ascaris constituting the highest risk. Since this infection is quite uncommon in Denmark, viruses and protozoa infections are of greater concern. A risk of 20%, as was calculated for scenario 1 when not taking incidence into account, may be considered too high for accepting the system with local use of faeces, and far exceeds the risk of 1:10 000 (10^-4) per year suggested as an acceptable level. By employing various measures to improve the situation it would however probably be possible to significantly reduce the risks from pathogens. It is difficult to obtain sufficiently high temperatures in small scale composting/dehydrating/storage systems even if other material such as straw or wood chips are added for structure. It is probably easier to add a pH-elevating material such as ash or lime to ensure an inactivation of the pathogens and a stabilization of the material. Exposure to material stored for less than 12 months should be avoided and even after this period exposure should be minimized.

Risks from pharmaceuticals and micro-pollutants

Environmental risk. Predicted environmental concentrations in soil were calculated for the 25 most sold pharmaceuticals in Denmark. For only three of these compounds - Oestradiol, Ibuprofen and Digoxin, effect data on environmental relevant species could be found. The risk quotient RQ defined as, RQ = PEC/PNEC, was calculated for all the scenarios presented above. For the three compounds the following range of RQs applied:

• Oestradiol [41.000- 376.000]
• Ibuprofen [0.3 - 13]
• Digoxin [0.005-0.5]

Results thus clearly show that for two out of the three compounds the PEC/PNEC ratio exceed the ratio of one, and therefore implies that the compound affects the terrestrial environment. It should however be emphasised that the evaluation has been performed on calculations and not using measured concentrations.

Lack of data for toxicity is the primary obstacle for assessing the environmental risk of pharmaceuticals. Nevertheless the results obtained, especially for Oestradiol, imply that this type of compound may affect the environment and suggest that much more effort should be made in order to produce the needed data for a more comprehensive risk analysis.

The levels in faeces of most of the evaluated heavy metals and dioxins were well below existing Danish quality criteria for soil and sludge intended for application on soil with the exception of cadmium and mercury. These metals may be present in faeces at levels about equal to or slightly exceeding the quality standards for use on agricultural soils.

Human health risk. Calculations showed that humans need to consume 200 g to 1 kg soil in order to be exposed to one adult Daily Defined Dose (DDD) of the drug. As stated above Larsen (1998) anticipated that it was not possible to eat more than 10 gram of soil per day. Thus, it is not possible for either children or adults to be exposed to a whole DDD on one day via soil. On the other hand it is difficult to establish a "safe level" for pharmaceuticals such as hormones, antibiotics and cancer drugs. A "zero tolerance" principle might therefore be proposed and applied to protect vulnerable groups of the human population. Some groups (e.g. children) have for different reasons enhanced sensitivity towards drugs.

The evaluation of human health risk for micro-pollutants was based on an exposure scenario where a 10 kg child accidentally ingests 200 mg faeces/day. This dose of faeces will not for any of the micro-pollutants evaluated lead to an exposure level higher than 1/16 of the TDI.

Conclusions. The most important conclusions from this project were:

• There is an unacceptably high risk of infection by pathogens when faeces is used locally without treatment.
• The highest risk was attributable to Ascaris. According to the calculations several secondary infections may be expected from each container with infectious material.
• The risk of infection from bacterial pathogens are low if the material is stored for 12 months unless the pathogens can regrow in the material.
• The risks of infection can be reduced by simple measures such as longer storage, or treatment with a pH elevating compound. This will have to be applied to all containers, as it will be impossible to identify only infected containers.
• For Oestradiol the PEC/PNEC ratio far exceeded the ratio of one, implying that the compound affects the terrestrial environment.
• Calculations showed that humans need to consume 200 g to 1 kg soil in order to be exposed to one adult Daily Defined Dosis (DDD) of a drug. A "zero tolerance" principle may however be applied to protect vulnerable groups in the population.
• Exposure to micro-pollutants during or after local handling of faeces does not appear to be a significant environmental or health problem.

| Forside | | Indhold | | Forrige | | Næste | | Top |

Version 1.0 Juli 2005, © Miljøstyrelsen.