回分式好氧污泥床操作方式為廢水注入時間2 min，曝氣時間231 min (氣體表面流速 0.0138~0.0277 m/s，溶氧維持5 mg/L以上)，沉澱時間5 min，排水時間2 min，一次回分操作時間為240 min，一天共6次回分操作。回分式好氧污泥床處理乙酸及葡萄糖基質在有機負荷率2、4、6、及8 kg COD/m3-d操作條件下，COD去除率皆達98%以上，且反應器中污泥顆粒化與沉降性皆良好。無論處理葡萄糖或乙酸基質，污泥顆粒平均粒徑及比攝氧率皆隨著有機負荷率之增加而增大，但污泥齡則隨著有機負荷率之增加而降低。處理葡萄糖基質反應器中之污泥濃度(7900~10950 mg VSS/L)及生物質量(30~41g)都略高於處理乙酸基質者(7900~10060 mg VSS/L；30~38g)，處理葡萄糖基質之污泥顆粒粒徑(1.09~1.94 mm)亦略大於處理乙酸基質者 (1.02~1.21 mm)，且處理葡萄糖基質之污泥顆粒比攝氧率(62~96 mg O2/g VSS-h)亦略大於處理乙酸基質者(53~88 mg O2/g VSS-h)。
經由獨立批次實驗求得葡萄糖好氧降解之動力常數intrinsic k (12.4~23.8 d-1)、Ks (81~90 mg COD/L)及apparent k' (8.9~13.9 d-1)、K's (100~137 mg COD/L)皆略大於乙酸好氧降解者(k值11.7~22.0 d-1、Ks值47~56 mg acetate/L、k'值8.3~11.9 d-1、K's值62~103 mg acetate/L)，而二者apparent k'值都小於intrinsic k值，且皆隨著有機負荷率之增加(污泥齡降低)而增大，apparent K's值則隨著有機負荷率之增加(污泥顆粒粒徑增大)而增大，並大於intrinsic Ks值，由處理乙酸基質之質傳參數(??2 = 41~100、Bi = 15~25及?? = 0.24~0.31)與處理葡萄糖基質之質傳參數(??2 = 160~1089、Bi = 22~44及?? = 0.11~0.16)得知，回分式好氧污泥床之污泥顆粒外部質傳阻抗對整體基質去除速率之影響不大，但污泥顆粒內部之質傳阻抗對整體基質去除速率之影響則相當大，又以葡萄糖基質之影響更大，且內部質傳阻抗隨著有機負荷率和污泥顆粒粒徑之增加而增大，回分式好氧污泥床中污泥顆粒之內部質傳速率將是整體基質去除速率之限制步驟。本研究建立之動力模式及經驗模式模擬之COD去除率與回分式好氧污泥床處理乙酸及葡萄糖基質實驗値之誤差在 3%範圍內，且動力模式與經驗模式模擬值之間差異百分比亦僅在1.42 %範圍內。
A kinetic model (incorporating intrinsic kinetics) and an empirical model (incorporating apparent kinetics) of the sequencing aerobic sludge blanket reactor (SASBR) treating non-inhibitory substrates are formulated. Meanwhile, four SASBRs were used to treat non-inhibitory substrate glucose and acetate, respectively, by maintaining at four different organic loading rates (OLRs). Thus, not only the performance of SASBRs treating non-inhibitory substrates can be evaluated but the associated mass transfer and reaction kinetics also be elucidated. The proposed kinetic and empirical models were validated by experiments as well.
Each SASBR was operated with a cycle length of 240 min, 6 cycles a day. One cycle consisted of 2 min of feeding, 231 min of aeration (superficial air velocity = 0.0138–0.0277 m/s, DO>5 mg/L), 5 min of settling, and 2 min of discharging. When the SASBRs were used to treat glucose and acetate, respectively, by maintaining at the OLRs of 2, 4, 6, and 8 kg COD/m3-d, not only the COD removal efficiency of greater than 98% can be reached, but fairly good sludge granulation/settling also be achieved. For respectively treating glucose and acetate, with an increase in OLR, the average granule diameter (dp) and the specific oxygen utilization rate (SOUR) increased, whereas solids retention time (SRT) decreased. Moreover, the biomass concentration and its quantity for treating glucose (7900–10950 mg VSS/L; 30–41 g) were slightly greater than those for treating acetate (7900–10060 mg VSS/L; 30–38 g). dp for treating glucose (1.09–1.94 mm) was slightly larger than that for treating acetate (1.02–1.21 mm). The SOUR of aerobic granules for treating glucose (62–96 mg O2/g VSS-h) was slightly higher than that for treating acetate (53–88 mg O2/g VSS-h).
From the independent batch experiments, the obtained intrinsic k, Ks (12.4–23.8 d-1; 81–90 mg COD/L) and apparent k', Ks' values (8.9–13.9 d-1; 100–137 mg COD/L) for treating glucose are all slightly greater than those for treating acetate (k = 11.7–22.0 d-1; Ks = 47–56 mg acetate/L; k' = 8.3–11.9 d-1; Ks' = 62–103 mg acetate/L). No matter glucose or acetate was treated, the apparent k' values are all smaller than the intrinsic k values; both the apparent k' and intrinsic k values increase with increasing OLR (decreasing SRT). Meanwhile, the apparent K's values increase with increasing OLR (increasing dp); the apparent K's values are significantly greater than the intrinsic Ks values. The calculated mass transfer parameter values (for acetate: ??2 = 41–100, Bi = 15–25, ?? = 0.24–0.31; for glucose: ??2 = 160–1089, Bi = 22–44, ?? = 0.11–0.16) show that the influence of internal mass transfer resistance (increases with increasing OLR and dp) on the overall substrate removal rate in the SASBRs is much greater than the influence of external mass transfer resistance. This implies that the internal mass transfer resistance can be regarded as rate-limiting. The calculated COD removal efficiencies using kinetic and empirical models are only 3% deviated from the experimental results. The variations of the simulated results using kinetic and empirical models are within 1.42%.