Sewage sludge: odour emissions assessment

6. Discussion of Results

6.1 Measurement Uncertainties & Other Variables

There are two main measurement uncertainties. The first factor is the repeatability of odour concentration determinations using olfactometric measurement with an odour panel. The precision limitations of olfactometric analysis data when interpreting monitoring data are set out in an annex to the BS EN 13725 standard. The 95 percent confidence limits for identical samples collected from the same source are set out in Table 2 below.

The data in Table 2 represents the limits of tolerable variations in measurement results for odour analysis in a laboratory complying with the BS EN 13725 standard and demonstrates that the 95th percentile range of "acceptable" results for a single sample covers a range of x 4.87 from the lower limit to the upper limit, whereas triplicate samples reduce this range to x 2.50. Long term practical experience is that such extremes in measurement results are extremely rare.

The results reported in Table 1 and Figure 7 above include 95 percent confidence limit values for the area specific odour emission rates based purely on these factors for triplicate samples, that is ignoring other variables, and also based on the simplistic assumption that all three samples collected from each sludge cake treatment were of identical odour concentrations. However, it is acknowledged that this may notbe entirely true if there were some progressive reductions in sludge cake surface emissions through even the short sampling periods required to collect three samples.

Table 2. The 95 percent confidence intervals for analysis of odour samples with an expected odour concentration of 1,000 ouE/m3 ( BS EN 13725 Annex F)

Number of Identical Samples Analysed	95% confidence limits for an expected odour concentration of 1000 ouE/m3
	Lower limit	Expected	Upper Limit
1	453	1,00	220
2	571	1,00	175
3	633	1,00	158
4	673	1,00	148
5	702	1,00	142
6	724	1,00	138
7	741	1,00	134
8	756	1,00	132

The other potentially significant measurement uncertainty concerns the ventilation rate of the Lindvall hood because of factors including the 'porosity of the cake and the seal between the sampling hood and the cake material, both of which may result in some leakage from the hood. External wind speed may also affect the rate of ventilation because of wind effects on the "layflat" tubing used as inert ducting. For the sake of consistency, all airflow measurements were made at the exit from the hood to reflect the actual ventilation rate of air through the entire length of the air flow path through the hood. It is estimated that these measurement uncertainties could account for differences of up to factors of x2 to x5 at the extremes of the measured emission rates.

It is also noted that the measurements reported here represent the sludge cake outputs from three example works. The composition of sludge cakes, and the resulting emission rates from cakes from other works may be quite different even with the same treatment processes.

6.2 Effects of Sludge Cake Treatments

After taking account of potential measurement uncertainties and errors, there were materially higher odour emissions from the limed cake than from the AD and THP treated cakes, with emissions in excess of an order of magnitude higher than the AD and THP digested cakes. These differences were also reflected in the indicative H2S emission rate measurements. The implications are that there are significantly higher risks of adverse odour emissions from land applications of limed cake than from applications of conventional AD or THP digested cakes. These findings are consistent with ADAS experience of measurements made during commercial consultancy work with limed and THP cakes elsewhere. The low odour emissions from AD and THP are perhaps not surprising because limed cake does not undergo the significant biological degradation which does occur in conventional anaerobic digestion, and particularly in the more destructive combination of THP and AD processes.

No H2S was detected in the THP digested cake and odour emissions were approximately half the rate of conventional AD cake, although these approximate factor of two differences may not be significant in relation to the measurement uncertainties.

Ammonia emissions were similar for the limed and THP digested cakes, but materially lower for the conventional AD cake. Elevated NH3 emissions could be expected with limed cake treatment because of the nature of the reactions caused within the cake by lime treatments. The nitrogen content of the pre- processed sludge cake is also likely to influence NH3 emissions from the resulting cakes.

6.3 Effects of Sludge Cake Age

The differences in odour emission rates between different sludge cake ages were not large in relation to the measurement uncertainties, although there was a trend for higher emissions from the most recently processed limed and conventional AD cake. Hydrogen sulphide emission rates for these two cakes were also higher from the most recently processed cake. This effect might be expected if there is a degree of additional biological or chemical stabilisation in cake which has been stockpiled for longer periods. On the other hand, NH3 emissions were slightly higher from the older limed and AD cakes than the more recently treated materials.

6.4 Context of Odour Emission Rates

The measured odour emission rates reported here can be compared with data from similar measurements made on other materials, including compost and other sewage-related materials.

In relation to composting and mature compost, the South West Industrial Crops Environmental Body^[1] sponsored a research project to measure odour emissions from composting process and compost. The findings were that odour emission rates from the surfaces of various feedstocks and stages in the composting process extend over a very wide range.

At one extreme odour emissions from predominantly green waste composting on one site, or from green waste composting with some wood and municipal feedstock based materials on a second site, were typically between 20 and 40 ouE/s/m2 and less than 20 ouE/s/m2 respectively. At a third site, where more

putrescible vegetable and plant materials were being composted, significantly higher emissions were measured on the feedstock materials and disturbed windrows/stockpiles, with odour emission rates up to 1,244 ouE/s/m2.

Higher, short-term emissions do occur during, and after turning with emission rates up to factors of 10 to

20 times higher during and after turning than from undisturbed windrows. Mature, well made compost was shown to be less odorous, with area specific emission rates comparable to those reported here from AD and THP cakes at around 2.5 ouE/s/m2. This data suggests that emission from land applications of AD and THP cake will be similar to those from stable/mature compost, and that emissions from limed cake are substantially higher than from compost.

In relation to sewage sludge and sludge cake the United Kingdom Water Industry Research (UKWIR) organisation has provided examples of "estimated" odour emission rates set out below in Table 3 in a technical guidance document2^[2].

Table 3. UKWIR Estimated Odour Emission Rates

Material & condition	Odour Emission Rates (ouE/s/m2)
	Low	Typical	High	Very high
Quiescent raw, liquid sludge tank or lagoon	7.9	40	80	160
Raw liquid sludge with some disturbance	140	710	1,400	2,800
Digested sludge tank (liquid)	14	71	280	1,400
Sludge cake	0.8 (old, digested)	62 (fresh, digested)	80 (raw)	800 (during disturbance)

The UKWIR "old, digested" sludge cake emission rate data is quite consistent with the measured emission rates reported here for AD and THP cake, but does suggest that fresh digested cake may be more odorous than the rates measured on the field stockpiles.

The substantially higher measured emission rates for LIMED cake reported here (242 to 347 ouE/s/m²) are at the upper end of UKWIR emission rates for sludge cake and comparable with emission rates for some

categories of liquid sludge.

6.5 Implications of results for the biosolids industry

This work concerns a small number of different sludge cakes from a limited number of different sewage treatment works, but on the evidence of the measured emissions rates the following conclusions can be drawn:

1. There are much higher risks of adverse odour effects from the land applications of limed cakes than there are from the application of anaerobically digested cakes, either with or without preliminary thermal hydrolysis.

2. The substantially higher odour emission rates from lime treated cake demonstrate that much more rigorous odour mitigation measures must be used than for digested cake (with or without preliminary THP) if odour impacts are to be controlled or mitigated. Examples of such additional controls could include:

a. Selecting application sites which are remote from residential settlements and housing,

b. Restricting applications to small areas of land at any one time,

c. Not applying limed cake to grassland or other areas which preclude ploughing-in or cultivation other than in very remote locations.

d. Ploughing- in or incorporating more or less immediately after land spreading so that only minimal areas of spread material are exposed between spreading and incorporation.