Sie sind auf Seite 1von 187
1634 Chapter 15 Issues Related to Teeatment Plant Performance Process Design Considerations for Solids Processing 1721 Odor Control 1723 PROBLEMS AND DISCUSSION TOPICS 1723 REFERENCES 1726 15-1 In the previous chapters of this book, special attention has been devoted to the unit oper- ations and chemical and biological unit processes used for the treatment of wastewater. While the importance of these subjects cannot be overstated, a number of other issues related to treatment-plant performance are of equal, if not of greater, impcrtance in the planning and design of wastewater management facilities. The two most important issues are the need to treat a broader range of constituents and the fact that discharge permit limits are becoming more stringent. In response to these issues, the following, topics are addressed in this chapter: (1) the need for upgrading treatment-plant per- formance, (2) treatment process reliability and the selection of design values, (3) odor ‘management, (4) process control strategies, (5) upgrading wastewater treatment plants, (© energy efficiency in wastewater treatment, and (7) considerations in the design of new wastewater treatment plants. NEED FOR UPGRADING TREATMENT-PLANT PERFORMANCE To meet both current and future wastewater discharge permit requirements and to meet more stringent requirements for discharge and reuse, most wastewater treatment facili- ties will need to undergo significant modifications and improvements. Meeting Current and Future Needs In the 20th century the primary focus of wastewater treatment was on the removal and treatment of settleable and floatable solids, organic matter expressed as biochemical ‘oxygen demand (BOD), total suspended solids (TSS), and pathogenic microorganisms (ee Chap. 1). Late in the century, nutrient removal and odors also became an issue. In the 2st century, wastewater treatment will still include the aforementioned con- stituents, but will also include a wide variety of human and veterinary antibiotics, human prescription and nonprescription drugs, industrial and household wastewater products, and sex and steroidal hormones (see Table 2-20 in Chap. 2). To treat many of these compounds, it will be necessary to optimize and upgrade existing treatment faci ities and operations. In some cases, additional treatment beyond secondary treatment, such as the-advanced treatment processes described in Chap. 11, will be required. For example, a considerable amount of upgrading will be required to meet new odor requirements at existing wastewater treatment plants. In some situations, new facilities will have to be designed and built. 15-1. Need for Upgrading Teeatment-Plant Performance 1635 Meeting More Stringent Discharge Requirements Based on past events, itis clear that the discharge requirements specified in NPDES permits will become more stringent in the 21st century. For example, the following con- stituent values were specified in a recently issued (2001) NPDES permit in California to be complied with by the year 2006. Limit Constituent Unit Type Value Type of sample BOD mg/L Monthly average 7 Daily composite: mg/L Weekly average 85 Daily composite mg/L Daily maximum 10 Daily composite Ts mg/L Monthly average 10 Daily compote mg/L Weekly average 18 Daily composite Triholomethanes (THMs) Bromodichloromethane g/L’ Daily maximum 1.14 Monthly grab ng/L “Monthly average 0.56 Monthly grab Chloroform ng/L Daily average 93 Monthly grab ng/L — Monthly average 5.7) Monthly grab Dibromochloromethane yig/L_ Daily average 0.92 Monthly grab Al Monthly grab ug/L Monthly average ‘Comparing the BOD and TSS values given above and the secondary standards given in Table 1-3 in Chap. 1, it is clear that treatment processes will have to be designed and operated differently to meet these standards. The total daily average for the three THMs is astounding, given that the current (2001) limit for total THMs in drinking water is 100 pg/L. Further, the constituent values given above are specified as not-to-exceed values, subject to fines and legal action for any exceedances. The State of California has enacted legislation (AB 307) by which fines can be levied for an indi- vidual exceedance, depending on the constituents and circumstances, and a mandatory fine is levied if more than four exceedances occur during any six-month period. Discharge Limits for Wastewater Treatment Plants Because not-to-exceed values are impossible to achieve with current wastewater treatment- plant facilities without massive expenditures, it is appropriate to define what is an acceptable level of performance. For example, 6 and 3 exceedances per year correspond to a 98.3 and 99.2 percent level of compliance, respectively. The U.S. EPA recommends an average frequency for excursions of both acute and chronic criteria not to exceed once in 3 years or 99.9 percent (U.S. EPA, 1994). It is interesting to note that not-to- exceed standards have been applied in England and Europe for the past 10 years; how- ever, the not-to-exceed limit is usually set at the 95 percent level. For values that exceed 95 percent, fines can be levied. With nondegradation groundwater requirements that are 1636 Chopler 15 15-2 {ssves Reloted to Teotmen-Plant Performance making their way into new NPDES permits, even 99.9 percent compliance may not be acceptable. ‘To meet permit limits, such as those given above, at the 98.3, 99.2, or 99.9 percent compliance level, it will be necessary to consider (1) the variability of the influent wastewater characteristics, (2) the inherent variability of biological treatment processes, and (3) the reliability of the mechanical treatment-plant components. Even with effec. tive design and operation, it may not be possible to meet very stringent discharge requirements at the 99.9 percent compliance level for a number of constituents without the use of additional treatment processes. Management of the additional residuals that ‘would result from enhanced treatment is another issue that must be addressed. The need for process optimization, additional treatment processes, or even the implementation of new treatment technologies will have to be addressed on a case-by-case basis, depend- ing on the constituent or constituents that must be removed and the levels to which those constituents must be removed. The question of what NPDES permit limits are set and whether these limits are appropriate is left to the courts and to another book. TREATMENT PROCESS RELIABILITY AND SELECTION OF DESIGN VALUES Reliability of a treatment plant or a treatment process may be defined as the probability of adequate performance for a specified period of time under specified conditions or, in terms of treatment-plant performance, the percent of the time that effluent concentrations meet specified permit requirements. For example, a treatment process with a reliability of 99 percent is expected to meet the performance requirements 99 percent of the time, For 1 percent of the time, or three to four times per year, the not-to-exceed daily permit limits would be expected to be exceeded. Such a level of performance may, or may not, be acceptable depending on the permit requirements. For each specific case where the reliability concept is to be employed, the levels of reliability must be evaluated includ- ing the cost of the facilities required to achieve the specified levels of reliability and the associated operating and maintenance costs. Thus, the purpose of this section is to exam- ine how treatment process variability is assessed and how the performance of combined processes can be evaluated. The specific topics to be discussed are (1) variability in wastewater treatment, (2) selection of protess design parameters, (3) the performance of combined processes, and (4) the development of input-output relationships. Variability in Wastewater Treatment ‘Three categories of variability that can affect the design, performance, and reliability of ‘a wastewater treatment plant are (1) variability of the influent wastewater flowrite and characteristics, (2) inherent variability in wastewater treatment processes, and (3) vari- ability caused by mechanical breakdown, design deficiencies, and operational failures. Each of these categories of variability is considered in the following discussion. Fol- lowing the discussion of these topics, an approach to the selection of design values in which the observed variability can be taken into account is introduced and illustrated ‘subsequently in Example 15-1. 15-2 Treatment Process Reliability ond Selection of Design Values 1637 ity of Influent Wastewater Flowrate and Characteristics. Because of the way modern human life is organized, there is variability in the influent wastewater flowrates and characteristics observed at wastewater treatment plants (see Chap. 3). A typical example of the variability that can be observed in influent waste- water flowrates and characteristics is illustrated on Fig. 15-1. As shown on Fig. 15~la, the summer flowrates are very stable and follow a log-normal distribution, whereas the entire daily set of flowrate data is influenced significantly by the high winter flowrates and is therefore extremely variable. In fact, the complete daily flowrate data set cannot bbe modeled with either an arithmetic or a log-normal distribution. As will be discussed subsequently, such variability, which is not uncommon, is of concen where stringent discharge requirements must be met. In some cases, it will be necessary to reduce the ‘amount of infiltration/inflow in the collection system and/or install equalization facili- ties to improve treatment process performance. The use of equalization facilities has the added benefit of reducing the size of the individual treatment units that may be needed, In general, as the capacity of the treatment facility increases, the observed variability in flowrates will tend to decrease. Inherent Variabi Wastewater Treatment Processes. Because of the variability of the influent wastewater flowrate and characteristics, the variability of all treatment processes due to design limitations, the inherent variability of biological Daily inftuent BOD data My = 245 m/l 8g Poa./Ps0 = Frais 112 Daily intuont TSS data Mg = 230 mit Cohcontraton BOD and TSS, mg/L 5 Peas/Peo 120 Py x : A nS st g Zao z i «| 20] 24 q ” MSee sees & EB Peron of values equal forse than nated value ® Figure’ 15-1 Probability distributions for doi © fluent wastewater characterises collected over‘one year: (a flowrate and {t} biochemical ‘oxygen demand (BOD) and total suspended solids (TSS). 1638 Chopler 15. Issues Reloted to Treotment-Alont Performance treatment processes due to the presence of living microorganisms and the laws of chance, all physical, chemical, and biological treatment processes exhibit some mea- sure of variability with respect to the performance that can be achieved. Based on an analysis of effluent performance data from a number of wastewater treatment processes, it has been observed that the performance of most processes for most constituents can be modeled with a log-normal distribution. In some cases, when dealing with advanced ‘wastewater treatment processes, many of the effluent constituent values will be reported as nondetect (ND), method detection limit (MDL), or level of quantification (LOQ) (see Sec. 2-1 in Chap. 2), When such values occur in a data set, the data set is said to be cen- sored in a statistical sense. Methods that can be used to estimate summary statistics for censored data sets are discussed in Helsel and Cohn (1998) and Helse! (1990). Typical performance summaries for various processes for a variety of constituents that have been presented in previous chapters of this textbook are summarized in Table 15-1. The variability in these distributions, measured in terms of the geometric standard deviation, 5, i8 also presented in Table 15-1 for the purpose of discussion and comparison. The ‘greater the numerical value of s,, the greater the observed range in the measured values. Table 15-1 Nariability observed in selected treatment processes taken from various figures presented in previous chapters Primary influent BOD an g/t 175 220 | 1.26 Primary influent TSS an g/t 150 195° 127 Effluent turbidity from granularmedivm filration ns. Nu 14 181.29 following biological treatment Effluent TSS from granulor-medivm filtration folowing 11-19 mg/L 1.95 28 14d biological treatment Effluent turbidiy from disk iter following biological © 11-31a._— NTU 0.65 085 1.31 freaknent Effient TSS from disk filer following biological treatment 11-31a mg/L 0.58 0.90 1.55 Primary influent coliphoge 19-16 phu/100m 21x 10' 35x 10 1.67 Secondary eluent coliphage following ocivated-shudge 13-16 phu/T00ml 300 5001.67 treatment Filler effvent coiphage following secondary 13-16 phy/100mt 140 4503.21 sedimentation Galiphage following disinfection with chlorine 13-16 phu/100 mt 2 5 250 TDS in reverse osmosis efvent following micrafitraion 13-170 mg/L 16 a 1.31 Seeded coliphage removal with reverse osmosis 13417 log removal 59 68 115 All of the reported distributions ore lognormal Pag = Paa/Po0- Table 15-2 Statistical measures used to assess equipment reliability? 15-2 Treatment Proces Rlibiliy and Selection of Design Values 1699 In general, it has been found that the performance of conventional biological treat- ment processes for wastewater constituents that can be modified by treatment, stich as BOD, TOC, etc., can be described most frequently with a log-normal distribution. For constituents that are not modified significantly such as inorganic TDS, both arithmetic and log-normal distributions can be used to model process performance. Where the variability in performance is not great, both the arithmetic and log-normal distributions can be used to model the observed performance. The Weibull distribution (Kokoska and Zwillinger, 2000) has also proven to be useful in the analysis of the performance of advanced wastewater treatment processes (WCPH, 1996, 1997). It should also be noted that censored effluent data sets are encountered often in assessing the performance of advanced wastewater treatment processes (see above), Mechanical Process Reliability. In addition to the variability in the influent wastewater flowrate and characteristics and the inherent variability in the response of wastewater treatment processes, the vetiability associated with the mechanical equip- ment used at wastewater treatment fatilities must also be considered in analyzing what design values and how much standby equipment must be available to meet stringent standards at some specified reliability value (e.g. 99 or 99.9 percent). Several approaches are available for analyzing mechanical reliability of a treatment plant including the fol- lowing (WCPH, 1996). 1. Critical component analysis (CCA) 2, Failure modes and effects analysis, 3. Event tree analysis 4, Fault tre analysis, All four of these approaches are cited frequently in the literature and are used by a vari- ety of industries. The critical component analysis (CCA) approach was developed by the U.S. EPA to determine the in-service reliability, maintainability, and operational availability of selected critical wastewater treatment components (U.S. EPA, 1982). The objective of the CCA is to determine which mechanical components in the wastewater treatment plant will have the most immediate impact on effluent quality should failure occur (WCPH, 1996, 1997). The statistical parameters used most commonly in apply- ing the CCA method, as defined in Table 15-2, are mean time before failure (MTBF), : egBesecioti ie EE. ‘Mean time before failure (MTBF) A measure of the mechanical reliability of equipment, Sein by the robert res. eo cpproach is divide the operting hours bythe nonber of flores oe Expected time before failure (ETBF) Similar to the MTBE, but he actol elapsed time i used {3 the total time in service Inherent aveilaility (AV Fraction of calendar time thet the component or unit : was operating Operating availbiliy (AVO) Fraction of time the component or unit can be expected to be operational excluding preventive mointenance “Adopted from U.S. EPA (1982), WCPH (1996, 1997) 1640 Chapter 15 Table 15-3 Summary statistics ‘on the mechanical reliability for ‘Agua lile® Issues Reloted fo Treatment Plant Performance Statistical measure 90% cL a MTBF, year MTBF, year. = AVO. avi Preliminary (headworks) 0.35 0.57 0.9953 0.9998 Primary 0.62 0.65 0.9967 0.9981 Secondary 242 175 09757 0.9953 Tertiary 224 178 0.9994 0.9995 UV disinfection 0.58 0.25 0.9991 0.9984 Reverse osmosis 122 0.99 0.9900 0.9903 Aeration tower 116 0.50 0.7835 0.9995 Carbon tower 1.86 1.02 0.9963 0.9999 Product water 0.56 0.45 09771 0.9964 Adopted from WCPH (1997). Aqua il dota 10/9/94 through 9/30/95, «See Table 15-11 for definitions of statistical measures. 490 percent confidence level expected time before failure (ETBF), inherent availability (AVI), and operating avail- ability (AVO). Ina study of the San Pasqual Aqua III treatment system in San Diego, CA, a com- plete process reliability analysis was performed (WCPH, 1996, 1997). The Aqua IIL treatment system was designed to produce 3800 m°/d (1.0 Mgal/d) of reclaimed water. ‘The treatment facility includes preliminary treatment (coarse screening and grit removal), primary treatment (rotary drum and disk screens), secondary treatment (water hyacinth ponds), and tertiary treatment (a package plant consisting of coagulation, soft- ening, sedimentation, and filtration). Advanced wastewater treatment consists of ultra- violet disinfection, reverse osmosis, air stripping, and granular activated carbon adsorp- tion. Sodium hypochlorite is used for plant effluent disinfection, with the required ccontact time taking place within the distribution system. The results of the process reli- ability analysis are presented in Table 15-3. As shown, the preliminary treatment process has the lowest MTBF. Typical problems experienced with the preliminary treat- ‘ment facilities included tripped breaker, packing leak, and gearbox failure. With the exception of three treatment processes, the AVO for the remaining processes was greater than 0.99. The AVI was greater than 0.99 for all of the treatment process (WCPH, 1996, 1997). Information such as that presented in Table 15-3 is used to deter- mine maintenance schedules and the requirements for standby parts and backup com- ponents. Selection of Process Design Parameters to Meet Discharge Permit Limits Because of the variations in effluent quality, treatment plants must now be designed to produce an average effluent concentration below the permit requirements. The question 15-2 Treatment Process Relioiliy nd Selection of Design Values 1641 is: What mean value should be used for process design to be assured that constituent concentrations in the effluent will be equal to or less than a specified limit with a spec- ified degree of reliability? Two approaches can be used to estimate the design mean value needed to meet prescribed standards: (1) a statistical approach involving the coef- ficient of reliability and (2) a graphical approach. Both of these approaches are described and illustrated in the following discussion. Whether the mean value arrived at can be designed for is a question that is addressed subsequently. Statistical Approach to Selection of Mean Design Value. One approach that can be used to determine the mean design value involves the use of the coefficient of reliability (COR) approach developed by Niku et al. (1979, 1981). In the COR ‘method, the mean constituent values (design values) are related to the standards that must be achieved on a probability basis. The mean value, m,, may be obtained by the relationship my = (COR)X, (154) where my = mean constituent design value (e.g., g/m?, mg/L) X, = a fixed standard that must be met at a specified reliability level (e.g., g/m’, mg/L) COR = coefficient of reliability, unitless ‘The coefficient of reliability is determined using the following expression. COR = [(V3 + 1)'PJexp {—Zs-elin (V3 + 1)!" 015-2) where V, = coefficient of variation of the ratio of the existing distributio (ee also Table 3-10 in chap. 3) = standard deviation of performance values (ie, daily, weekly, or monthly) from an existing treatment process mean of performance values (i.e., daily, weekly,or monthly) from an existing treatment process jumber of standard deviations away from mean of a normal distribution ‘@ = cumulative probability of occurrence (reliability level) om, ‘Values of Z,_.. for various cumulative probability levels, 1 ~ a, are given in Table 15-4. ‘Values of COR for determining effluent concentrations for different coefficients of vari- ation at different levels of reliability are reported in Table 15-5. Selection of an appro- priate value of V, must be based on data from operating facilities. The use of the relia~ bility concept is illustrated in Example 15~1, following the discussion of the graphical approach to the selection of process design values. Graphical Approach to Selection of Mean Design Value. Another method of determining appropriate mean design values to meet tife specified effluent standards is the graphical probability method. If it is assumed that the geometric stan- dard deviation can be used as a measure of reliability and that the value remains approx- imately constant at other design values, then the required éffluent value can be set at the specified reliability level (e.g., 10 mg/L at 99 percent) and a line is passed through the value with the same geometric standard deviation as the measured data. The value at a probability value of 50 percent is the new mean design value. 1642 Chapter 15. Issves Related to Treatment-Plant Performance Table 15-4 ‘Cumulative probability "Percentile Values of ive pre alg standardized normal distribution? 9.9 3.090 9 2.326 8 2.054 95 1.645 2 1.405 90 1.282 80 70 60 50 0.842 0.525 0.253 ° - From Niky ef ol. (1979) 03 1 081 O71 069 064 057 053 0.42 04 108 078 066 063 057 049 O44 0.33 05 112 075 061 058 O51 042 037 0.26 06 117° 073 087 054 047 0.37 0320.21 07 122 072 054 050 043 033 0.28 0.17 08 4128 O71 052 048 040 030 025 0.15 09 «135 «070 050 «0.46 «038 = 028 «0222 10 1.41070 0490.44 036 0280200. 12-156 070-046 «0.41033 0.22017, 0.08 15 180 070 045 039 030 019 014 0.06 ‘From Niku et al. (1979). ‘The graphical approach is illustrated on Fig. 15-2. The plotted data correspond to the monthly total copper concentration values in the effluent from a wastewater treat- ‘ment plant. Ifthe permit limit is to be 10 ug/L at the 99.9 percent reliability value, that value is plotted and a line with the same geometric standard deviation is drawn through the point. The value at a probability of 50 percent on the line drawn through permit limit corresponds to the required mean design value, in this case’2.1 g/L. For many con- stituents, it will be found that the required mean design value cannot be met with an existing process. Where the required mean value cannot be met with a single process, Figure 15-2 Probabily dstibutions: {o} monthly effluent ttl copper (Cu) concentration values collected over a period of 2 years ond (8 corresponding clstibution fo ochieve 99.9 percent com with e copper permit iit OF 10 ug/L drown with the some geometric stondord deviction asthe “xiginal dstibution 15-2 Treatment Process Reliobilly and Selection of Design Values 1643 Cu= 100, pol st 99.9% 10 “otal copper, uot (Gu design mean = 2.1 uot ° go 7 SRS RRESS B 099 Percent of values equal to orloss than indicated value it may be necessary to use two or more treatment processes in series. The typical geo- metric standard deviation values given in Table 15-1 can also be used for the design of, new treatment plants if the discharge permit limits are known. ‘An existing activated sludge plant is required to be expanded W permit requirements. The new effluent requirements are as 1644 Chapter 15 Issues Related to Treatment lant Performance ‘Monthly performance data BOD, —Tss, Month mg/L mg/L Jon 340 150 . 71 18.0 29.0 Ws 250 25 25.1 220 220 249 27 280 205 25.1 és 19.5 20.0 20.1 215. ae iven data using a standard statistical software 15-2 Treatment Process Reliability and Selection of Design Volues 1645, Determine the value of V, using the results ofthe statistical analysis i, For BOD ‘The value of Z,.. for a cumulative probability of 99% from Table 15-4 is 2.326. c. Determine the coefficient of reliability. i, For BOD IR = [ (0.194? + 1)? ] exp {— 2.326{in (0.194? + 1)]!”"} = 0.69 ii, For TSS : COR = [(0.172? + 1)?]exp {= 2.326[In (0.1727 + 1)]!7} = 0.71 3. Determine the eoeticient of reliability using Eq sae ata cumulative pro Dility of 99.9%. Determine the value of V, from Step 2a i. For BOD V, = 0.194 For TSS ‘V, = 0.172 a). Mean design BOD, = (COR) X, = 0.69 X 15 mg/L = 10.4 mg/L 'b. Mean design TSS = 0.71°X 15 mg/L = 10.7 mg/L Determine aed effluent concentrations for 99.9% reliability for the weekly 50g i 59X20 mg/L = 11,8 mg/L e i oer 26mg 1646 Chapter 15. Issues Related to Treatment Plant Performance 6. Use the most conservative values for design. BOD aign = 10.4 mg/L TSSeign = 10.7 mg/L Solution Part 2 Log-Probability Graphical Method 1. Plot the monthly data for BOD and TSS on log-probability paper. The required plots for BOD and TSS are shown below. eo 35 50} 50] ‘Weekly standard" eomge a 93 ley « “ 2 B alos : E a g* 3 i | 20} | Monty é mgt LPs] é (LR al ‘0 ‘0 "| H L at +—} 70 righ. | AA J05 mgt. H it H ter Lee = eee oe i a oae 8 BESS. 8 a8 5 2 a8 ical touadeaia ‘ten tas ego corlees than indicated valve orlaes than Inciated value 2. Estimate the design effluent concentrations for BOD and TSS for (a) 99% relia ‘Dility for the monthly standard and (b) 99.9% reliability for | . a. Determine the design effluent concentrations for BOD. ‘The BOD concentra- “~ “tions are determined by passing lines with the same slope as the measured data through the points at 15 mg/L and 99% and 20 mg/L and 99.9% and not- ing the conesponding values at 50%. The vals so determined ar: BODjui at 15 mg/L and 99% = 10.0 mg/L 4 BODacign at 20 mg/L and 99.9% = 11.0 mg/L b._ Determine the design effluent concentrations for TSS. The TSS concentrations are determined by passing lines with the same slope as the measured data through the pointsat 15 mg/L. and 99% and 20 mg/L and 99.9% and noting the “corresponding values at 50%."The values so determined are; 15-2 Treatment Process Reliability and Selection of Design Values 1647 TSS gain at 15 mg/L and 99% = 10.5 mg/L TSS gig at 20 mg/L and 99.9% = 11 mg/L It is interesting to note that the values determined graphically are essentially the same as those determined analytically using the COR method. Comment When the concept of reliability is used, the mean effluent values selected for design will typically be significantly lower than permit requirements. Loge et al. (2001) demon- strated how the coefficient of variation was reduced (je. process variability) when processes are used in series with respect to the disinfection of treated effluent with UV radiation. Based on numerous past designs in England and Europe, it has been found that to achieve the prescribed limit at the 95 percent level the average design value should be about 50 percent of the prescribed limit. Seana ene nnd RMR Scion Performance of Combined Processes Applying the statistical and graphical procedures as illustrated in Example 15-1, it will often be found that the resulting mean design value for a given process is well below the range where any factual knowledge exists on how to design the process. For exam- ple, assume that to meet NPDES permit requirements, an activated sludge process must be designed to meet an average effluent suspended solids concentration of 4 mg/L. Realistically, it is now not possible to design a secondary clarifier to meet a specific design value. What is typically assumed is that with good design and effective opera- tion of the secondary process, an average TSS value of 4 or 5 mg/L can be achieved. Unfortunately, such assumptions are not acceptable when not-to-exceed permit limits must be met. In such a situation, it will usually be necessary to add an additional process, such as depth or surface filtration, to meet the permit requirements consis- tently. In the following example the basis for determining the performance of the com- bined treatment processes is addressed. i, ase he following date apply to the activated sludge process. ‘Also, assume that n0 c ‘hemicals ‘are to. be usec with the depth filter. 1. Distribution for effluent 185 faoe ‘bern © 2. Geometric mean for effluent TSS, My "15 1 Me 3, pene ao Hie for TSS, 1648 Chopler 15 Issues Related to Treatment Plant Performance when the effluent turbidity is 2 NTU, the influent turbidity is 7.5 NTU, and the intercept of the line of best fit passes through 0.5 NTU. For these conditions the corresponding relationship is: Filter effluent turbidity, NTU = 0.5 NTU + 0.2(filter influent turbidity, NTU) To determine the TSS values after filtration, use the following two relationships given previously in Chap. 11. a. Secondary effluent TSS, mg/L = (2.3 mg/L/NTU) (effluent turbidity, NTU) (Eq. (11-32)] b. Filter effluent TSS = (1.4 mg/J/NTU) (filter effluent turbidity, NTU) {q. (11-33)] Using the above relationships, determine the values corresponding to 50% and 84.1% (one standard deviation) after filtration a. At 50% Secondary effluent turbidity = (15 mg/L)/[(2.3 mg/LYNTU] = 6.52 NTU Filter effluent turbidity = 0.5 NTU + 0.2(6.52 NTU) = 1.8 NTU iii, Filter effluent TSS = [(1.4 mg/LY/NTU] (1.8 NTU) = 2.52 mg/L b. At 84.1% [use Eq. (3-9) from Chap. 3] i. Pag = Pap XS, = 15 mg/L X 1.25 = 18.75 mg/L fi, Secondary effluent turbidity = (18.75 mg/LY{(2.3 mg/LYNTU] = 8.15 NTU iii, Filter effluent turbidity = 0.5 NTU + 0.2(8.15 NTU) = 2.13 NTU iv. Filter effluent TSS = [(1.4 mg/L)/NTU] (2.13 NTU) = 2.98 mg/L_ Plot the secondary effluent TSS and the filter effluent turbidity and TSS and pre- pare a summary compliance table. a. Log-probability plot a5 Fever is ein Sateen 15-2 Treatment Process Reliobilty and Selection of Design Values 1649 b. Summary table Value at indicated proba! Item Unit 98.3% 99.2% 99.9% 1s. mg/L 26 29 42 Turbidity NTU 25 28 3.0 Comment As shown in the summary table, 6 TSS exceedances per year (98.3 percent probability) would be equal to or greater than 2.6 mg/L, 3 TSS exceedances per year (99.2 percent probability) would be equal to or greater than 2.9 mg/L, and 1 TSS exceedance in 3 years (99.9 percent probability) would be equal to or greater than 4.2 mg/L. If a tur- bidity level of 2 NTU had to be met for reuse applications, the 2 NTU limit would be exceeded 25 percent of the time, without the use of additional treatment. snake Anaemia einai Development of Input-Output Data To predict the performance of different combined processes, input-output relationships such as presented in Example 15-2 will have to be developed. In general, input-output relationships are needed for processes with short treatment retention times such as gran- ular and surface filtration, microfiltration, ultrafiltration, and reverse osmosis. Input- ‘output relationships for processes where the treatment process performance is decou- pled from the hydraulic retention time, such as the activated sludge process, are not very useful because of the buffering provided by the process. Further, as illustrated in Sec. 7-S in Chap. 7, the value of the effluent substrate concentration for the activated sludge process is independent of the influent concentration. The variability observed in the effluent BOD, TSS, and other constituent values from activated sludge processes reflects (1) the inherent variability of the biological process (notwithstanding conclu- ms reached from the theoretical equations); (2) the diurnal, daily, monthly, and sea- sonal variation in the influent flowrates and wastewater characteristics; (3) the impact of process adjustments made by operators; and (4) the configuration of the physical facilities. For most processes with short retention times, the output values will be affected by the influent concentrations. It is interesting to note that the performance of the cloth media disk filter, as illustrated on Fig. 11-310 in Chap. 11, is essentially inde- pendent of the influent turbidity for turbidity values up to 30 NTU. ‘A further and very important consideration in the development of input-output rela- tionships for processes following the activated sludge process is the solids retention ’ time (SRT) at which the process is operated. As illustrated on Fig. 12-48 in Chap. 12, the number of particles with one or more associated coliform organisms will vary with the SRT. It has also been observed that the performance of downstream processes employing some form of filtration will also vary with the SRT value. Thus, the SRT value must be taken into account when developing input-output relationships for down- stream processes. In general, it will not be possible to use input-output relationships developed for processes following activated sludge processes with long SRT values 1650 Chopter 15) 15-3 Issues Related to TreotmentPlant Performance (© 104) to predict the combined performance of downstream processes follo. vated sludge processes with short SRT values (1 to 2 d). ODOR MANAGEMENT The potential release of odors is a major concern of the public relative to modifying existing wastewater treatment facilities and constructing new facilities. Thus, the con- trol of odors has become a major consideration in the design and operation of waste- water collection, treatment, and disposal facilities, especially with respect to the public acceptance of these facilities. In many instances, projects have been rejected because of the fear of potential odors. In several states, wastewater management agencies are now subject to fines and other legal action over odor violations. In view of the importance of odors in the field of wastewater management, the following discussion will dea! with (I) the types of odors encountered, (2) the sources of odors, (3) the movement of odor- ‘ous gases, (4) strategies for odor control, (5) odor control methods, and (6) the design of odor-control facilities. ‘ Types of Odors For humans, the importance of odors at low concentrations is related primarily to the psychological stress the odors cause, rather than to the harm they do to the body. The principal types of odors encountered in wastewater management facilities are reported in Table 15-6. With few exceptions, odorous compounds typically contain either sulfur or nitrogen. The characteristic odor of organic compounds containing sulfur is that of decayed organic material. Of the odorous compounds reported in Table 15-6, the rot- ten egg smell of hydrogen sulfide is the odor encountered most commonly in waste- water management facilities. As noted in Chap. 2, gas chromatography has been used successfully for the identification of specific compounds responsible for odors. Unfor- tunately, this technique has not proved as successful in the detection and quantification of odors derived from wastewater collection, treatment, and disposal facilities, because of the many compounds that may be involved. It should be noted that at higher con- centrations, many of the odorous gases (e.g., hydrogen sulfide) can, depending on expo- sure, be lethal. Sources of Odors ‘The principal sources of odors in wastewater management facilities and the relative potential for release of odor are presented in Table 15-7. Minimization of odors from these sources is the concern of odor management, Wastewater Collection Systems. The potential for odor release from.collec- tion systems is high. The principal sources of odorous compounds in collection systems are from (1) the biological conversion, under anaerobic conditions, of organic matter containing nitrogen and sulfur, and (2) the discharge of industrial wastewater that may contain odorous Compounds or compounds that may react with compounds in the ‘wastewater to produce odorous compounds. Odorous gases released to the sewer atmo- sphere can accumulate and be released at air release valves, cleanouts, access ports (i.e. ‘manholes), and house vents.

Das könnte Ihnen auch gefallen