Modified University of Cape Town (UCT) process


Modified UCT process is virtually same as Virginia Initiative Plant (VIP) process, but typically UCT process is designed for the SRT of 13-25 days while UCT process is designed for 5-10days. But, when it comes to MBR, the difference between the two processes does not exist. The major benefit of UCT process is it does not directly return the nitrate bearing mixed liquor to anaerobic tank, thereby ORP remains low enough to maximize phosphate release from phosphorus accumulation organisms (PAO). However, the cascade type mixed liquor recycle to anaerobic tank causes diluted MLSS in anaerobic tank. To compensate the MLSS dilution in anaerobic tank, tank size needs to be increased.

In modified UCT process, similar to MLE process, mixed liquor in rear end of the process is recycled to the front end of the process using a cascade type mixed liquor recycle as shown in Fig. 1, where the mixed liquor in membrane tank is recycled to aeration tank, the mixed liquor in aeration tank is to anoxic tank, and the mixed liquor in anoxic tank is to anaerobic tank. The internal recycles, QM/O, QO/AO, and QAO/AA range 2Q-5Q, 1Q-2Q, and 1Q-2Q, respectively, depending on condition and the goals in nutrient removal.


Fig. 1. Process diagram of University of Cape Town (UCT) process applied for MBR.


Effective HRT in each tank


One thing noticeable is that due to the high recycle rates, effective HRT in the tanks are much lower than the apparent HRT calculated based on the reactor volume and the wastewater flow rate. If median recycle rates are taken (QM/O=3.5Q, QO/AO= 1.5Q, and QAO/AA=1.5Q), actual HRT in anaerobic, anoxic, and aerobic tanks are only 1/3, 1/4, and 1/4 of their apparent HRT. For instance, apparent HRT in anoxic tank is 1-3 hours, but the effective HRT can be only 0.17-0.50 hour due to the mixed liquor recycle. Under this short HRT condition, flow short circuiting can play a significant role especially when mixing power is not sufficient. Therefore extra care must be taken to allow extra mixing in the place mixed liquor is introduced. Table 1 summarizes recycle ratio of each stream, apparent HRT, and effective HRT.


Table 1. Effective HRT in each tank of UCT process. Average values are in parentheses.

Flow Destination Anaerobic Anoxic Aerobic Membrane
Flow Origin Feed 1Q
Anaerobic 2Q-4Q (3Q)
Anoxic 1Q-3Q (2Q) 2Q-6Q (4Q)
Aerobic 1Q-5Q (3Q) 3Q-6Q (4.5Q)
Membrane 2Q-5Q (3.5Q)
Apparent HRT (hrs) 0.5-1.0 1-3 3-5 1-2
Effective HRT based on average flow (hrs) 0.17-0.33 0.17-0.50 0.40-0.67 0.22-0.44

MLSS and effective SRT in each tank

The relative MLSS in each tank of modified UCT process can be calculated by performing the mass balance around each tank. The four assumptions used in this calculation are as follow.

  • MLSS produced in each tank when raw wastewater passes through the system is negligible comparing to the existing MLSS
  • MLSS in each tank is in steady state
  • MLSS in raw wastewater is negligible comparing to the MLSS in anaerobic tank
  • The flow rate of excess biosolids removal from the membrane tank is negligible compared with QM/O.

In the following calculations, Q indicates the raw wastewater flow rate and X indicates MLSS.


Input    =   Output   

Aerobic tank      :                 image002123        ——— (1)

                Anoxic tank        :      image0043456    ——— (2)

                Membrane tank :        image0067890      ——— (3)

The solutions of the above equations are:

image00812qw ———————- (4)


image01034er    ———————- (5)


image012qw45             ———————- (6)

For example, if QAO/AA, QO/AO, and QM/O are assumed at 2Q, 2Q, and 4Q, respectively, the relative MLSS among the four tanks are XAA : XAO : XO : XM = 0.36 : 0.53 : 0.80 : 1.00. It is apparent that there are significant MLSS imbalance among the tanks, which ultimately affects the nutrient removal efficiency by affecting the exposure time of microorganisms in each different environment.

The effective solids residence time (SRT) in each tank can be calculated using the following equation, where qx is mixed liquor removal rate from membrane tank and xm is MLSS in the same tank.

image01812345asd              ———————- (7)


In one study, the performance of the UCT process was compared with that of Sammamish Biological Nutrient Removal (SmBNR) process in lab scale (Johannessen, 2006). The SmBNR process is identical to HANT process, but it does not have deoxygenation tank and the internal recycle rate is set at very high (5Q). In this study, the modified UCT process had a combined aeration and membrane tank. The QM/AO was set at 5Q, which was much higher than the typical (1Q-2Q), while QAO/AA was set at a typical level, 2Q. The major difference between the two systems is that fresh wastewater is fed to anaerobic tank in UCT process while it is fed to anoxic tank in SmBNR process. Table 2 summarizes the performance of the two processes. Due to the opposite arrangement of anaerobic and anoxic tanks in the two processes, UCT was better in P removal while SmBNR was better in nitrogen removal.

Table 2. Performance comparison of UCT and SmBNR processes

Parameter Influent (mg/L) UCT Effluent (mg/L) SmBNR Effluent (mg/L)
COD 747 11.6 11.0
NH4-N 57.9 0.3 0.3
NO3-N 0 16.1 10.1
PO4-P 20.0 6.6 11.4
bCOD/P 35 46



The relative MLSS in each tank in modified UCT process can be calculated using the following spreadsheet calculator. The membrane tank MLSS in output can be manually modified.



© Seong Hoon Yoon