1、1180 单词， 6600 英文字符， 1975 汉字 Controlling sludge settleability in the oxidation ditch process K.J. Hartley ABSTRACT This paper describes an investigation aimed at developing an operating technique for controlling sludge settleability in the oxidation ditch form of the nitrification denitrification act
2、ivated sludge process. It was hypothesized that specific sludge volume index (SSVI) is lowest at an optimum process anoxic fraction and increases at higher and lower fractions. Using effluent ammonia:nitrate ratio as a surrogate for anoxic fraction, it was found that a simple empirical model based o
3、n a three solids retention time moving average nitrogen ratio was able to replicate the long-term SSVI variations in two independent oxidation ditches at a full-scale plant. Operating data from a second oxidation ditch plant during periods when a prefermenter was on- or off-line showed that SSVI als
4、o varies with RBCOD, greater RBCOD giving lower SSVI. It was concluded that best settleability occurs at about the same anoxic fraction as lowest effluent total nitrogen concentration, with an ammonia:nitrate ratio of about 1. An operating rule of thumb is to use dissolved oxygen control to maintain
5、 effluent ammonia and nitrate nitrogen concentrations about equal. A third oxidation ditch plant deliberately operated in this manner achieved 15-month median operating values for SSVI of 60 mL/g and for effluent ammonia, nitrate and total N, respectively, of 0.2, 0.3 and 2.0mgN/L. Keywords Settleab
6、ility ； Sludge volume index； SVI； Control； Nitrogen； Oxidation ditch 1 Introduction The oxidation ditch process has the capability to reliably meet low total nitrogen standards because of its special operating characteristics. Nitrogen removal can be continuously optimized in operation by varying th
7、e anoxic and aerobic mass fractions to meet short- and long-term changes in wastewater characteristics (Hartley, 1997； Mines and Woods, 1994). It has also been observed that sludge settleability in oxidation ditches varies with the operating mass fractions (Vivian and Hartley, 1989). Sludge settleab
8、ility is one of the most important design and operating characteristics of the activated sludge process. It affects both the sizing of tankage and the process capacity actually achieved. Yet, nearly 100 years after the activated sludge process was invented (Ardern and Lockett, 1914), positive contro
9、l of this parameter remains elusive. Four slightly different tests are in use for measurement of settleability, sludge volume index (SVI, employing simple cylinder settling, Mohlman, 1934), stirred SVI (sSVI, in which the cylinder is fitted with a stirrer, APHA et al, 1998), diluted SVI (DSVI, using
10、 serial dilutions of the sludge, Koopman and Cadee, 1983) and SSVI3.5, or simply SSVI (stirred specific volume index, in which the sludge concentration is standardized at 3.5 g/L, White, 1975). The values of these parameters can be related to the settling properties of the sludge, and to each other,
11、 through published statistical correlations (Daigger, 1995； Ekama et al., 1997； Ozinsky and Ekama, 1995). A range of factors affects settleability in activated sludge processes, including feed quality, solids retention time (SRT), mixing characteristics, oxygen regime, pH and nutrient availability (
12、Jenkins et al., 2004). For N and N&P removal processes, the Water Research Group at the University of Cape Town has conducted a comprehensive settleability study, which they published in a lengthy series of papers and summarised in Ekama et al. (1996), Casey et al. (1999) and Tsai et al. (2003). The
13、ir study included intermittent aeration formats mimicking the oxidation ditch process. They found that the low F/M group of filaments (such as Microthrix parvicella and 0092) cause practically all bulking problems in long SRT nutrient removal processes and that they cannot be controlled by anaerobic
14、, anoxic or aerobic selectors. Whether or not the process includes biological phosphorus removal, it is the operating nitrogen regime that governs sludge settleability. According to the model they developed, in a two-stage anoxicaerobic system utilising mainly slowly biodegradable COD, if denitrific
15、ation in the primary anoxic zone is incomplete and nitrite is present, the intracellular intermediate nitric oxide persists and inhibits the denitrifying floc-formers in the aerobic zone. The filamentous organisms, which do not compete well in the anoxic zone because they can only denitrify nitrate
16、to nitrite, are not inhibited in the aerobic zone where they proliferate. Intracellular nitric oxide does not accumulate when the COD is readily biodegradable (RBCOD) so a significant fraction of RBCOD relieves the competitive pressure on the floc-formers. Tsai et al. (2003) observed that M. parvice
17、lla growth could be stimulated or inhibited by varying the concentration of ammonia (the organisms required nitrogen source) to high or low values, respectively. They therefore postulated that deterioration of settleability at low aerobic fraction can be due to increasing ammonia concentration (decr
18、easing nitrification efficiency). 2 The oxidation ditch Fig. 1 illustrates the general oxidation ditch reactor format. Mixed liquor is impelled around a circuit by one or more mechanical aerators, or by mixers in the case of diffused air aeration. This provides a typical mixed liquor: feed flow rati
19、o of 100200. The dissolved oxygen (DO) concentration jumps up at the aerator and declines downstream. The DO concentration at the aerator equals the oxygen transfer rate divided by the circulating flow rate. The anoxic fraction is controlled by varying the aeration rate to maintain a DO set point at
20、 a probe located towards the downstream end of the aerobic zone. The anoxic fraction can be varied by adjusting the set point (Hartley, 1997). Fig. 2 shows the process-operating characteristic. There is an anoxic fraction-operating window that gives lowest total nitrogen (TN) in the effluent. This w
21、indow shifts as operating conditions (feed and process operating characteristics) change. Lowest TN occurs where effluent ammonia and nitrate are about equal. The process can therefore be optimised in operation to maintain lowest (or any other) TN concentration. Because the mixed liquor recycle rati
22、o is so high, the ammonia and nitrate concentrations are almost constant, and equal to the effluent concentrations, throughout the circuit. According to the UCT model, sludge settleability in an oxidation ditch should be best when TN is lowest, deteriorating at a lower anoxic fraction where the oxid
23、ised nitrogen concentration exiting the anoxic zone is elevated, and also at a higher anoxic fraction where the ammonia concentration increases. Actual SVIs in operating ditches are variable. For example, von Munch and Komarowski (2001) report an unstirred SVI value of 105 mL/g for the Thorneside pl
24、ant and an sSVI value of 155 for the Gibson Island plant (both in Brisbane, Australia). At the Bundamba plant in the same region the unstirred SVI has been variable between 90 and 400 mL/g (Vivian and Hartley, 1989). Two Florida plants, Dale Mabry and Falkenburg, were reported to operate at unstirre
25、d SVIs of 140 and 180 mL/g, respectively (Hartley, 1995； Mines and Woods, 1994). This paper analyses the operating data from three Australian oxidation ditch plants to demonstrate that (a) the settleability characteristic does behave as expected, (b) the UCT model appears to require modification to
26、explain the observations and (c) operating settleability can be readily controlled in the oxidation ditch process. The three plants are Bucasia (located in Mackay), Coolum and West Byron (Byron Bay), all treating domestic sewage. Physical and operating data are summarised in Table 1. Routine operating data have been used for the analysis and no information is available on the types of filamentous organisms present.
