To investigate the interactive effect of biosolid treatment and post treatment dewatering techniques on biosolid total N, inorganic N, total C, and C:N ratio, five biosolid/sludge types that have undergone various treatment and post treatment drying techniques were selected (Table 1). Similarly to evaluate the effect of biosolid/sludge treatment method alone on the N content of biosolid/sludge, municipal wastewater sludge of the same origin was exposed to different treatment processes (activated sludge without digestion [Activated] and anaerobically digested [ADS3]). Then after both Activated and ADS3 were dried on concrete beds in thin layers of about 100 mm. These two products were selected because they are the most commonly produced byproducts in South African WWTP. Of the five biosolids, three of them (ADS2, Activated, and Thermally Hydrolysed Sludge [THS]), were selected for laboratory N mineralization study to determine the interactive effect of biosolid treatment and post treatment dewatering techniques on the N fertilizer value of biosolids. The effect of drying thickness/depth on the N content of biosolid was investigated using liquid sludge collected after mesophilic anaerobic digestion (ADS2L).

Representative samples for each sludge/biosolid treatment and drying technique were collected randomly from the drying beds for analyses. The samples were analysed in triplicates. The actual sewage treatment processes and dewatering/drying methods evaluated in this experimental work are shown in Table 1 while selected chemical properties of the sludge materials are presented in Table 2.

A laboratory N mineralization study was conducted using three sludge/biosolid types (ADS2, Activated, and THS) in a constant room temperature of 25 ±1 °C. Two of the sludge/biosolids (ADS2 and Activated) represent the most commonly produced sludges/biosolids in South Africa. The third, THS, is one of the technologies, which South African WWTP are considering to adopt. The laboratory N mineralization study was conducted in 1,163 cm³ volume airtight containers where 100 g soil was mixed with the candidate sludge/biosolid to meet a 300 kg N ha⁻¹ application rate.

The amount of biosolid/sludge added for each biosolid/sludge type under investigation to satisfy the 300 kg N ha⁻¹ is presented on Table 3. The soil used for the laboratory N mineralization study was a Hutton sandy clay loam (loamy, kaolinitic, mesic, Typic Eutrustox). Selected properties of the soil used for the laboratory N mineralization study are presented on Table 4. Inorganic N in the form of potassium nitrate (KNO₃) was added to increase the soil NO₃^--N level to 60 mg kg⁻¹ to minimize N immobilization [19] (Table 3). The soil moisture was maintained at field capacity by adding water based on mass difference.

Samples from the laboratory N mineralization study were collected for analyses at the following day intervals: 0, 1, 3, 7, 15, 30, 65, and 100. For statistical analysis purposes, each treatment had three replications for each sampling period. All treatments were aerated during sampling for the first seven days of laboratory N mineralization. After day seven, treatments were opened and weighed once a week and at sampling time to replenish water to field capacity.

To investigate the effect of drying thickness/depth on beds on the N content of biosolids, Anaerobically Digested Sludge (ADS2L) was used. This sludge was selected because it is the most commonly produced biosolid in South Africa used in agricultural lands. Selected chemical properties of the ADS2L biosolid are presented in Table 5.

Field experiment was conducted at one of East Rand Water Care Works (ERWAT) plants, Vlakplaats at Vosloorus, Sedibeng district, Gauteng. To mimic the drying bed system used at WWTP, plastic containers (0.62 m long × 0.62 m wide × 0.38 m deep) with drainage holes at the bottom were used. The basement of the drying container was then lined with a geo-fabric material that only allows liquid to drain whilst retaining all the solids. The plastic containers were dug into the ground allowing the brim of the drying container to be at level with the ground surface. Five drying depths (5, 10, 15, 20 and 25 cm), replicated three times, were then laid randomly in the drying containers. The study was conducted for two seasons (autumn [rainy and warm temperature] and winter [cooler and no rainfall]). The drying period for each phase was determined by the maximum number of days (drying time) required to reach 90-95% solid or 5-10% moisture content in drying beds.

During sludge drying, samples were collected in triplicates from the ‘drying beds’ for analysis of targeted physical and chemical properties. Sampling was done on weekly basis for the first three weeks and thereafter, in two weeks interval. Sampling continued from day 1 until biosolid solid concentration reaches 90-95% solid or 10-5% moisture content at which the experiment was terminated. During each sampling period, biosolid samples were collected from each treatment using a scoop and were placed into a plastic container and sealed. The plastic containers were transferred immediately into a mobile cooler maintained at 4 °C and were transported to the University of Pretoria soil laboratory for analyses.

Biosolid samples were ground to pass through a 150 m screen and analyzed for total C and N using Dumas combustion on a Carlo Erba NA1500 N, C, S elemental analyzer (Carlo ErbaStrumentazione, Milan, Italy). Organic C was determined using the standard Walkley-Black method [20]. Electrical Conductivity (EC) and acidity (pH) of the biosolid materials were determined in 1:2.5 biosolid to water ratio. Total P and all heavy metals were determined using Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES) (Spectroflame Modula, Spectro, Kleve Germany) after wet acid digestion [21].

Inorganic N (NO₃-N and NH₄-N) release from sludge amended soils and control were determined by the steam distillation method described by Mulvaney [22].

The cumulative amount of net N mineralized from biosolids was calculated using eq. (1):

$\text{Net N }\text{}\text{mineralized = Inorganic N }\text{}\left(\text{sludge amended soil}\right)-\text{ Inorganic N }\left(\text{control}\right)$ (1)

The Net N mineralized was converted to Net N mineralized per organic C applied [g kg⁻¹] and Net N mineralized per organic N applied [g kg⁻¹] using eq. (2) and eq. (3), respectively:

$\text{Net N }\text{}\text{mineralized }\text{}\left(\text{per organic C applied}\right)\text{ = }\left(\frac{\text{Net N mineralized}}{\text{Organic C}\text{}\text{ added from biosolid per kg soil}}\right)$ (2)

$\text{Net N }\text{}\text{mineralized }\text{}\left(\text{per organic N applied}\right)\text{ = }\left(\frac{\text{Net N mineralized}}{\text{Organic N}\text{}\text{ added from biosolid per kg soil}}\right)$ (3)

Data were analysed using ANOVA and the General Linear Model Procedures of Windows SAS Version 9.3 [23]. Error bars on figures representing the effect of biosolid treatment and post treatment drying techniques on the N content and N mineralization indicate standard deviation at P≤ 0.05.

Total N content differed across biosolid treatment and post treatment dewatering (drying) techniques (Table 6). Generally, biosolid total N content varied between 2.81% for ADS1 dried in thick layers of 250 mm to 5.47% for ADS3 dried in thin layers of about 100 mm (Table 6). The total N content of the anaerobically digested biosolids, which were dried in thick layers of 250 mm in two different drying bed types (concrete beds) ADS1 (2.81%) and (sand paddy) ADS2 (2.83%) were similar. Similarly, the total N content of anaerobically digested biosolid dried in thin layers of 100 mm (ADS3 [5.47]) was relatively closer to the N content of an activated sludge (Activated [4.95%]) dried in thin layers of 100 mm, despite the differences in treatment processes.

The similarity in the total N content between ADS1 and ADS2 as well as between ADS3 and Activated seems to indicate that drying depths and time play a crucial role in affecting the N content of sludges. Nonetheless, further study on the effect of drying depth on the N content of sludge was warranted. Hence a separate field trial was conducted to ascertain this findings (presented in the following section).

The total N content of THS (3.13%) was within the ranges of the total N contents of ADS1 and ADS2. The N content of both ADS1 and ADS2 were within the ranges reported by Barbarick and Ippolito [24] (1.6-3.6%) for biosolids that have undergone similar treatment and drying technique. The drying depth used to dry the biosolids reported by Barbarick and Ippolito [24], however, is not stated.

In contrast to the total N, the proportion of inorganic N (NH₄⁺ + NO₃^- + NO₂^-) was higher for biosolids dried in thick layers of 250 mm (ADS1 [11.2%] and ADS2 [9.7%]) than those dried in thin layers of 100 mm (Activated [2.4%] and ADS3 [2.7%]). The inorganic N proportion results from this study for ADS1 and ADS2 were within the ranges reported for biosolids from similar treatment processes (mesophilic anaerobic digestion) Cogger et al. [12] (9.3%) as well as Barbarick and Ippolito [24] (16.5%). The biosolid drying depths on beds for the latter were not reported. The inorganic N contents of ADS3 and Activated were, however, far lower than those reported by Rigby et al. [25] for biosolids from similar treatment processes.

The initial N content and C:N ratio are the basic properties that highlight the potential decomposition rate of a biological organic material [26]. Previous studies with plant litter have shown that materials with a low C:N ratio are characterized by rapid decomposition and less immobilization than materials with higher C:N ratios [27]. It is also well documented that organic materials which are rich in lignin, cellulose and hemicellulose are characterized by low decomposition rates [28].

Cumulative net N mineralization varied significantly among biosolid types (Figure 1). This is in agreement with previous studies which reported that N mineralization from biosolids is a function of biosolid treatment processes [9]. Generally net N mineralization from the anaerobically digested sludges (ADS2) (23 g N kg⁻¹ organic C applied) was higher than mineralization from similar biosolid types reported by Parnaudeau et al. [19] (< 10 g N kg⁻¹ organic C applied). This is most probably because of the relatively higher initial N content (2.83%) (Table 3) and lower C:N ratio (8.76) (Table 3) of the biosolids from this study compared with that of Parnaudeau et al. [19] (2.48% N, 9.19 C:N ratio).

Figure 1

N mineralization per kg organic C (a) and N mineralization per kg organic N (b) applied from anaerobically digested sludge dried in paddy in thick layers of 250 mm (ADS2), Activated sludge dried in thin layers of less than 100 mm (Activated) and THS

Net N mineralization per kg organic C (Figure 1a) and per kg organic N applied (Figure 1b) after 100 days of N mineralization study was higher from activated sludge than from the other two biosolids (ADS2, and THS). This was most probably attributed to the higher initial N content of activated sludge (4.95%) compared with that of ADS2 (2.83%) and THS (3.13%) (Table 2). Previous studies [29] have shown that the initial N concentration influences the amount of N that can potentially be mineralized. In addition, Activated sludge had lowest lignin fraction (3.21%) by mass compared with the ADS2 (15.96%) and THS (8.7%) (Figure 2). This is attributed to a combination of factors including the reduction of the mass of solid waste [30] and the accumulation of non-degradable lignin [29] with anaerobic digestion of the activated sludge. Lignin affects N mineralization negatively [28]. This has a significant implication on the N fertilizer value of the wastewater sludges.

Figure 2

Total percentage distribution of organic constituents for five sludge types (anaerobically digested dried in concrete beds in layers of 250 mm depth [ADS1], anaerobically digested dried in paddy in layers of 250 mm depth [ADS2], anaerobically digested dried on concrete slab in layers of less than 100 mm depth [ADS3], activated sludge dried in concrete slab in layers of less than 100 mm depth [Activated], THS) (Lip = Lipids, SOL = Soluble Organic Compounds, Hemi = Hemicellulose, Cell = Cellulose, Lign = Lignified fraction)

The N fertilizer value of activated sludge investigated in this study was 22 kg N per tonne of biosolid. This is in contrast to the N fertilizer value of THS (10 kg N per tonne biosolid) and ADS2 (6 kg N per tonne biosolid). Despite only a 10% difference in the initial total N content between THS (3.13%) and ADS2 (2.83%), cumulative N mineralization from THS (57.11 g N kg⁻¹ C applied) on day 100 was 2.5 times higher than that of ADS2 (22.27% g N kg⁻¹ C applied). Such a difference is most probably attributed to the higher lignin fraction (15.96%), lignin:N ratio (1.40), and C:N ratio (8.78) of ADS2 compared with THS (lignin = 8.7%, lignin:N ratio = 0.50, C:N ratio = 5.77). It is well documented that higher lignin:N ratio results in low net N mineralization as a result of N immobilization [19].

Generally, the total N content of the wastewater sludge showed an increasing trend during the drying period (Figure 3a and Figure 3b). The only exception is in autumn between week 3 and 5 (drying depths of 15, 20 and 25 cm) and between week 7 and 9 (drying depth of 25 cm). This is in contrast to the pattern of total C, which decreased as the time for drying progressed (Figure 3c and Figure 3d). The total N content of the sludge for the shallow drying depths of 5 cm and 10 cm (in autumn) and 5-15 cm (in winter) increased rapidly during the first two weeks of drying. At the end of the drying period, highest total N was recorded for the 10 cm drying depth both during the autumn (3.53%) and winter (3.72%) seasons (Figure 4). On the other hand, the lowest total N was recorded for the drying depth of 25 cm both during autumn (2.62%) and winter (2.65%) (Figure 4). Total N content of each drying depth remained similar between seasons. The only exception was for the 15 cm depth, where the N content for winter being significantly lower than autumn.

Figure 3

Total N in autumn (a); total N in winter (b); total C in autumn (c) and total C in winter (d) changes at 5, 10, 15, 20 and 25 cm drying depths during drying on drying beds (error bars indicate standard deviation)

Total N increased as biosolid drying time progressed thus leading to higher final total N of the dried biosolid than the initial liquid sludge, on dry mass basis. This finding agrees with that of Nayak et al. [31], who reported an increase in total N as time progressed during sludge compositing.

Figure 4

Total N content (oven dry mass basis) of anaerobically digested sludge dried at five drying depths (error bars indicate standard deviation)

The highest increment was observed for the 10 cm drying depth, followed by 15 cm and 5 cm drying depths. While thicker drying depths of 20 and 25 cm resulted in lower N by the end of drying period. The lower total N content recorded at 20 and 25 cm depths are most likely due to two reasons: enhanced NH₃ volatilization and denitrification losses from the middle to lower sections of the drying beds, which remained quite wet for longer time. The relatively wet and warm environment in the middle to the lower sections of the drying beds (data not presented) speeded up organic matter decomposition during which NH₄⁺ is released. An increase in the concentration of NH₄⁺, which was observed between weeks 3 and 7 in autumn and weeks 2 and 5 in winter for the 20 and 25 cm drying depths (data not presented), leads to an increase in the concentration gradient between the atmospheric NH₃ and the NH₃ in the biosolid resulting in the diffusion of ammonia from the biosolid strata to the atmosphere thus negatively affecting the total N content of the biosolid [17].