**8. Effect of height of the bioreactor on performance**

**6. Scalability of the process**

26 Nuclear Material Performance

The sulphate reduction was monitored under anaerobic condition in a 9-litre bioreactor for 25 days, and the percentage of mean reduction was 56.3 ± 11.03. In case of 78-litre bioreactor, the sulphate reduction was monitored for 361 days, and the percentage of average reduction was 62.36 ± 24.81 with the saturation obtained at around 331 days. The probability distribution of the data was found to be normally distributed. It was evident from the above observations that the efficiency of sulphate reduction was increased by 1.10-fold with scaling up of the bioreactor volume. Our results were quite similar to the study of Sarti et al. [28], where they had also successfully demonstrated sulphate reduction in an anaerobic bioreactor with maximum efficiency of 99%. A similar study of fed batch bioreactor used for sulphate reduction was demonstrated by Silva et al. in the year 2002 with an efficiency of 97% [29]. Although both the studies are similar to ours in terms of reduction efficiency, but unlike those, our system

The sulphate reduction was further optimized in terms of incubation time in the same bioreactor under identical conditions. The desired level of reduction (for environmental discharge) was reached within three and half hours (**Figure 7**). The data were fitted using

23 45 *Y x xx xx* =+ - + - + 0.941 160.51 182.51 83.8 18.14 1.39 (1)

= 0.97. The

Origin8pro (**Figure 8**), which resulted in a polynomial equation of order 5 with *r* <sup>2</sup>

retains the reduction efficiency consistently once it gets stabilized.

**7. Optimization of the time of sulphate reduction**

following equation perfectly expresses this desulfurization system:

**Figure 7.** Time course of sulphate reduction by the packed bed bioreactor.

The oscillatory nature of the bioreactor performance in terms of sulphate reduction was further analysed by calculating the running mean of the sulphate reduction of the samples taken from each port of the bioreactor as described in **Figure 4**. The result indicated that significant amount of sulphate reduction was observed between the first and second ports (18 cm), in contrast, there was no significant reduction between the second and third ports (40 cm; **Figure 9**). As an explanation to this observation, it can be argued that the compromised performance in the upper layer could be due to the accumulation of hydrogen sulphide gas generated by the system, which has a negative impact on system performance. The phenomenon was also supported by the report of Frank et al. in the year 2013 [30]. To decrease the dead space (where the performance is compromised) and enhance the efficiency of the system, an alternative reactor design was tested (**Figure 10**). Similar designs for one vertical system and one hori‐ zontal system were constructed and tested (**Figure 10b**). The performance of the two systems was found to be similar with no significant statistical variation observed using *z* test for the equality of two means with unknown variances and moderate sample sizes (*n* = 47 in case of both vertical and horizontal designs). The results are displayed in **Table 1**.

**Figure 9.** Left panel shows the position of the ports on the bioreactor column. The right top panel shows actual sul‐ phate reduction at the different ports, whereas the right bottom panel shows running mean of sulphate reduction from three different ports of the bioreactor for 60 hours. Bottom curve for the first port, middle one for the second port and the upper curve was for the third port.

Developing Tailor-Made Microbial Consortium for Effluent Remediation http://dx.doi.org/10.5772/62594 29

**Figure 10.** (a) Modified design of packed bed bioreactor. (b) Actual bioreactors constructed using plastic material.

**Figure 9.** Left panel shows the position of the ports on the bioreactor column. The right top panel shows actual sul‐ phate reduction at the different ports, whereas the right bottom panel shows running mean of sulphate reduction from three different ports of the bioreactor for 60 hours. Bottom curve for the first port, middle one for the second port and

the upper curve was for the third port.

28 Nuclear Material Performance


**Table 1.** Statistical validation of sulphate reduction using different designs of the bioreactor.

From the above observation, it was clear that the diminished performance in the upper layer was not due to accumulated hydrogen sulphide gas. There might be other factors responsible for this performance variation. As the bioreactor design was proper for the current system under investigation, the optimization of process was done using response surface methodol‐ ogy (RSM) under ambient condition and implemented using design expert 9 software as displayed in **Table 2**. Experimental and predicted responses were found to be broadly similar.


**Table 2.** Table representing the experimental design for system optimization using response surface methodology.

From the above analysis, the optimum sulphate reduction condition was determined at an initial sulphate concentration of 1250 ppm and at a flow rate of 1.8 litre/hour (**Figure 11**). The mathematical equation derived from the model is given below. The values of each term are given in the coefficient table (**Table 3**).

$$\text{Equation for sulphur reduction} = 49.07 + 2.61 \times A - 1.69 \times B - 0.28 \times AB + 8.02 \times A^2 + 2.56 \times B^2 \tag{2}$$

**Figure 11.** The model graph for sulphate reduction in response to sulphate concentration and flow rate.


**Table 3.** Statistical validation of the optimization study.

where *A* is the sulphate concentration, and *B* is the flow rate.

#### **9. Conclusion**

**Factor 1 Factor 2 Response 1 Predicted**

Sulphate reduction (%) Sulphate reduction (%)

B: flow rate (litre/hour)

**Table 2.** Table representing the experimental design for system optimization using response surface methodology.

From the above analysis, the optimum sulphate reduction condition was determined at an initial sulphate concentration of 1250 ppm and at a flow rate of 1.8 litre/hour (**Figure 11**). The mathematical equation derived from the model is given below. The values of each term are

2 2 Equation for sulphate reduction 49.07 2.61 1.69 0.28 8.02 2.56 = + ´- ´- ´ + ´ + ´ *A B AB A B* (2)

**Figure 11.** The model graph for sulphate reduction in response to sulphate concentration and flow rate.

 1 719.67 2.82 59.66 55.62 2 1780.33 2.82 66.05 60.3 3 1250.00 3.00 46.54 56.57 4 719.67 1.98 60.72 58.46 5 1250.00 2.40 49.07 50.63 6 1250.00 2.40 49.07 50.63 7 1780.33 1.98 68.21 60.3 8 1250.00 2.40 49.07 50.63 9 1250.00 2.40 49.07 50.63 10 1250.00 2.40 49.07 50.63 11 500.00 2.40 58.63 54.05 12 1250.00 1.80 53.85 52.93 13 2000.00 2.40 63.6 58.79

Standard Run A: sulphate

30 Nuclear Material Performance

concentration (ppm)

given in the coefficient table (**Table 3**).

The work contained in this chapter describes a biofilm-based soluble sulphate reduction system operating within 3.5 hours using a well-characterized SRB consortium from 1600 ppm to discharge level under ambient condition. This ensures the treatment of 1509 litres of sulphate solution in 24 hours using a 220-litre bioreactor. A single-unit bioreactor would be the ideal configuration for this consortium. Time kinetics of sulphate reduction yielded a parabolic form significantly (*r* <sup>2</sup> = 0.99; *p* < 0.05). Rate of sulphate reduction was found to be independent of seasonal variation. The bioreactor designs tested during this study had practically no effect on the performance of the system. This system was the fastest sulphate-reducing system at pilot scale, which could run without maintenance for a long time with the ability to withstand an initial sulphate concentration of 1250 ppm at a flow rate of 1.8 litre/hour optimally under ambient condition. Hence, the process has been filed as an Indian patent and a PCT to protect the intellectual property associated with this invention. It has immense application for industrial effluent treatment in future.

#### **Acknowledgements**

The authors would like to acknowledge the financial assistance of Ministry of Human Resource Development (MHRD), Government of India (GOI) under the FAST scheme for conducting part of the work; Department of Atomic Energy, GOI for initiating the work and scaling it up to 220 litres; Department of Biotechnology, GOI for providing fellowship to Poulami Datta, Shashi Bhushan, Ganesh Prasath Krishnan and Swati Bhatt; Department of Science and Technology, GOI under the DST Inspire Scheme for providing fellowship to Sourav Ghosh and MHRD for the fellowship of Chaitali Chanda. The authors would like to thank Late Sourav Chakratorty, Arpan Pal and Abhishek Mitra for their technical assistance. The authors would like to thank Dr. Gauri G. Pandit and Dr. Tessy Vincent of Bhabha Atomic Research Centre for their intellectual inputs during the execution of the project.
