Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 314
Watershed Management for Potable Water Supply: Assessing the New York City Strategy 8 Phosphorus Management Policies, Antidegradation, and Other Management Approaches This chapter discusses four policy tools that New York City is using, or could be using, within its watershed management strategy. Although somewhat unrelated to one another, these tools represent various ways in which protection strategies (both structural and nonstructural) can be implemented. In addition, they each highlight the difficulty of assessing and mitigating nonpoint source pollution in comparison to point source pollution. The Total Maximum Daily Load (TMDL) Program, mandated for impaired waters by the Clean Water Act (CWA), is a powerful tool for determining the relative contributions of point and nonpoint source pollution to the water supply reservoirs. The Phosphorus Offset Pilot Program, or "trading" program, was created to allow new point sources of pollution without increasing the overall level of pollution within a subwatershed. Antidegradation refers to a state and federal policy that is intended to prevent the lowering of water quality within water supply reservoirs (and other waterbodies). Finally, treatment processes beyond chlorine disinfection are being explored for use in the New York City water supply system. TOTAL MAXIMUM DAILY LOAD PROGRAM Section 303(d) of the CWA requires states to identify waters that do not meet the goal of "fishable, swimmable water quality" and to develop TMDLs for them. The TMDL process involves identifying the chemical(s) of concern that are causing impairment, defining a water quality standard for that chemical (if a federal or state standard does not already exist), and determining the allowable loading (TMDL) of that chemical such that the standard is not exceeded. If the current pollutant loading to a waterbody exceeds the TMDL, then the state must
OCR for page 315
Watershed Management for Potable Water Supply: Assessing the New York City Strategy identify point and nonpoint sources of pollution and suggest ways to decrease loading from these sources. Participation of all 19 New York City reservoirs in the TMDL program is required under the City's filtration avoidance determination. The New York City Department of Environmental Protection (NYC DEP) has been developing TMDLs for each reservoir in three phases; the TMDLs generated at each phase supersede the previous values and reflect more site-specific data, better modeling efforts, and improved implementation methods for meeting TMDLs. Because the overall goal of the TMDL program is to identify specific polluting areas and land uses and concentrate point and nonpoint source pollution controls to those specific areas, the TMDL calculations are viewed by NYC DEP as a planning exercise that will guide the implementation of best management practices (BMPs) throughout the New York City watershed (K. Kane, NYC DEP, personal communication, 1998). This fact places tremendous importance on the adequacy of the methods and models used in the TMDL program. NYC DEP has decided to base its TMDL calculations on total phosphorus as the compound of concern. In doing so, it has assumed that phosphorus is the limiting nutrient for plant primary production and that it correlates best with disinfection byproduct (DBP) formation, nuisance algae and eutrophication, hypolimnetic anoxia, and taste and odor problems. Using phosphorus as the target of the TMDL program presents several challenges because it is somewhat ambiguous and controversial as an indicator of drinking water quality. Because there is no national maximum contaminant level (MCL) for phosphorus, NYC DEP has had to interpret qualitative state laws regarding phosphorus. The state currently endorses a guidance value for in-reservoir total phosphorus of 20 µg/L measured in the epilimnion during the growing season (NYS DEC, 1993). This value, based on aesthetic effects for primary and secondary contact recreation, was used for Phase I TMDLs. For Phase II, however, NYC DEP has interpreted the state's narrative standard for phosphorus, which states that, for all classes, "there shall be none in amounts that result in the growth of weeds, algae, and slimes that will impair the waters for their best uses." (NYCRR 701-703). Accordingly, a phosphorus concentration of 15 µg/L has been recommended by NYC DEP and used for Phase II TMDLs (NYC DEP, 1999a). TMDL Methodology The TMDL program in New York City was developed in phases in order to generate results quickly with currently available information and also to update TMDLs on a regular basis as more data and sophisticated models become available. The first two phases have been completed; Phase I values have been adopted by the New York State Department of Environmental Conservation (NYS DEC) and approved by the Environmental Protection Agency (EPA), and Phase
OCR for page 316
Watershed Management for Potable Water Supply: Assessing the New York City Strategy II values await approval in September 1999. Phase III is not expected to be complete until 2002. Phase I Methodology Step 1: Calculating the TMDL. The first step of the TMDL process determines the status of the basin with respect to phosphorus concentration. Phosphorus concentration data were collected from different points in each reservoir and at different depths, both biweekly and monthly. During Phase I, data from the growing seasons (May 1 through October 31) of 1990–1994 were collected, and the median value for each year was calculated. Median values of phosphorus concentration were converted into annual phosphorus loads using the Vollenweider equation (see Equation 8-1 of Box 8-1). The Vollenweider equation is a simple, steady-state model of chemical flux through a completely mixed waterbody (Vollenweider, 1976). It was originally developed with data collected from Canadian lakes. The derivation of the Vollenweider equation and its use to convert phosphorus concentrations to phosphorus loads is discussed in Box 8-1. It should be noted that the use of the Vollenweider equation to estimate loadings to the water supply reservoirs is unusual. In most cases, the Vollenweider equation is utilized to estimate the total phosphorus concentration in the reservoirs based on measured or estimated loadings from tributaries and other inputs. Agreement between the measured and calculated phosphorus concentrations indicates appropriate application of the model. In NYC DEP's application of the Vollenweider equation, measured total phosphorus concentration in the reservoirs is utilized to "back-calculate" the loading to each reservoir (and a comparison is made with an independent method, the Reckhow equation, to estimate loadings). This use of the Vollenweider equation requires less information to be gathered because monitoring of storm events from tributary inputs to derive accurate estimates of nutrient loadings is not needed. The estimated phosphorus loads (L) from each of the five years between 1990 and 1994 were arithmetically averaged to determine one value for the "current load." The state guidance value for phosphorus, 20 µg/L, was also converted into an annual load (the "critical load") using the Vollenweider equation. The TMDL for a reservoir corresponds to this critical load. However, given the uncertainties inherent in the process, a margin of safety (10 percent for the New York City reservoirs) is usually subtracted from the critical load to give the "available load." It is this available load to which the current load is compared to determine if a waterbody is exceeding its TMDL. If the current load is below the available load, then the basin is not exceeding its TMDL. If the current load is greater than the available load, then the amount in exceedance must be reduced in the watershed by reducing the contribution of phosphorus from either point or nonpoint sources.
OCR for page 317
Watershed Management for Potable Water Supply: Assessing the New York City Strategy BOX 8-1 The Vollenweider Equation The Vollenweider equation was derived as follows: a simple mass balance for a constituent like total phosphorus in a complete-mix basin (in which a fraction of material is settling at an apparent settling velocity of vs) results in the following steady-state equation: where P = the total P concentration throughout the reservoir, mg/L L = the total P loading to the reservoir from all sources, g/m2/day qs = the overflow rate of the reservoir (flow rate/surface area), m/day vs = the apparent mean settling velocity of total P, m/day This mass balance equation can be written in an equivalent form as: where τw = the mean hydraulic detention time of the reservoir, days R = the fraction of total P retained within the reservoir system (dimensionless) H = the mean depth of the reservoir (volume/surface area) The form of the equation that the NYC DEP utilized varies from Equation 8-2 because it makes use of correlations to account for the fraction of total phosphorus that is not retained in the reservoir. The fraction of total phosphorus that passes through the reservoir is inversely proportional to the square root of the mean hydraulic residence time (Larsen et al., 1976). This empirical relationship, which has been validated on hundreds of lakes, was found to be accurate for the New York City reservoirs (Janus, 1989). Therefore, the final equation that is used by the NYC DEP is:
OCR for page 318
Watershed Management for Potable Water Supply: Assessing the New York City Strategy Step 2: Modeling All Sources of Pollution. The second step in the Phase I TMDL process is to establish the relative contribution of phosphorus from different sources throughout the basin. In New York City, this step used data from 1993 only. To determine the contribution from wastewater treatment plants (WWTPs), effluent samples were collected on a regular basis and a flow-weighted mean phosphorus concentration was determined. This concentration was then multiplied by the average flow from the WWTP and the number of days in operation during 1993 to determine the annual phosphorus loading from WWTPs. Upstream reservoirs are also considered to be point sources for downstream reservoirs. To determine their contribution, the tunnels connecting the reservoirs were sampled for phosphorus at a point just prior to the receiving reservoir. Nonpoint sources pose greater difficulties because few methods have been developed to measure their contributions. NYC DEP used the Reckhow model to predict the contributions of agricultural land, forest, urban land, atmospheric deposition, and septic tanks (Reckhow et al., 1980). As shown in Box 8-2, each contribution was calculated by multiplying the area of land devoted to that land use by an export coefficient, which was derived from the literature. For septic systems, only those sited within 100 ft of lakes, reservoirs, and watercourses were assumed to contribute loadings to the reservoirs. Because current state regulations prohibit septic systems within 100 ft of a waterbody, the septic system contribution is made up primarily of preexisting and noncomplying systems. The Reckhow equation sums up all the point and nonpoint sources and produces an annual phosphorus loading in kg/year, which can be converted into phosphorus concentration, again using the Vollenweider equation. This concentration was compared to the measured annual median phosphorus concentration. If the difference between those values was less than 20 percent, then the Reckhow model was assumed to be an adequate model for that basin. Although this second step may appear to be strictly academic and not required during the TMDL process, NYC DEP was obliged (because of the filtration avoidance determination and agreements with NYS DEC) to measure point and nonpoint phosphorus loading using the Reckhow model, even for those basins that were not exceeding their TMDLs. Step 3: Determining Wasteload Allocations and Load Allocations. The last step of the Phase I TMDL process was only done if a basin exceeded its TMDL (as it did in Cannonsville Reservoir). In this step, the waste load allocations (WLA) from point sources and load allocations (LA) from nonpoint sources are determined to help optimize management strategies. NYC DEP determined the Cannonsville WLA by assuming that all WWTPs would meet the effluent phosphorus concentration goals of the Memorandum of Agreement (MOA) as a result of the upgrades. These effluent phosphorus concentrations were then multiplied by the State Pollution Discharge Elimination System (SPDES) permit-
OCR for page 319
Watershed Management for Potable Water Supply: Assessing the New York City Strategy BOX 8-2 The Reckhow Model The Reckhow model is a lumped-parameter model that determines the contribution of different land uses to overall pollutant loads. Data on land use in the New York City watersheds are stored in a Geographic Information System (GIS) database derived from Landsat images taken at 30-m resolution. Four land use categories are distinguished in the Reckhow model: urban, forest, agriculture, and water, with the agricultural land use being sometimes subdivided into corn/alfalfa, bare soil, and grass/shrubs. For Phase I, the export coefficients for the Reckhow model were derived mainly from the literature. Slight variations in the Phase I coefficients were used to reflect known differences between the Croton, Catskill, and Delaware watersheds. In Phase II, only the Croton watershed TMDLs utilized the Reckhow model. Export coefficients for this phase were revised to include some site-specific data and information. Coefficients used for Phases I and II are shown below along with the governing equations for the Reckhow model. where L = the total P loading to the reservoir from all sources, kg/year ECag = export coefficient for agricultural land (kg/ha/yr) ECf = export coefficient for forest land (kg/ha/yr) ECu = export coefficient for urban land (kg/ha/yr) ECa = export coefficient for atmospheric input (kg/ha/yr) ECs = export coefficient for septic systems (kg/capita/yr) Aag = area of agricultural land (ha) Af = area of forest land (ha) Au = area of urban land (ha) As = area of lake surface (ha) PSI = point source input (kg/yr) Septic = septic system input (kg/yr) SR = soil retention coefficient. Export Coefficients (kg/ha/yr): Croton Phase I Catskill Delaware Croton Phase II Urban 0.70 0.70 0.70 0.90 Forest 0.05 0.07 0.05 0.05 Atmospheric 0.53 0.53 0.53 0.10 Agriculture: 0.3 0.3 Corn/alfalfa 2.00 2.00 Bare soil 0.30 0.30 Grass shrubs 0.15 0.20
OCR for page 320
Watershed Management for Potable Water Supply: Assessing the New York City Strategy ted flows to obtain the WLA. The LA was obtained by subtracting the WLA from the available load. The main purpose of calculating a WLA is to set a permanent upper bound on phosphorus loading that can be incorporated into the SPDES permit of a WWTP. The LA is sometimes used to target best management practices to the appropriate nonpoint sources. Phase II Methodology Phase II has a more complex methodology than Phase I, although the three steps of the process are analogous (NYC DEP, 1999b). Details are only provided where differences between Phases I and II exist. Step 1. In Phase II, the geometric mean of the reservoir phosphorus concentration data was used rather than the median. The geometric mean was thought to be a more accurate representation of the phosphorus concentration data, which follow a lognormal frequency distribution (K. Kane, NYC DEP, personal communication, 1998). Values that were below the detection limit (2–5 µg/L) were set to half the detection limit. Unlike Phase I, phosphorus concentration data from 1992 to 1996 were used in Phase II. The geometric mean from each of these five years was arithmetically averaged to determine the current load. The margin of safety in Phase II was changed slightly to account for large variations in phosphorus data from year to year (K. Kane, NYC DEP, personal communication, 1998). The margin of safety can vary between 10 percent and 20 percent, depending on interannual variability in phosphorus concentrations (NYC DEP, 1999b). Finally, the Vollenweider equation was altered slightly for Phase II to accommodate its coupling to the Generalized Watershed Loading Function (GWLF) in Step 2. Unlike the Reckhow model, which does not vary over time, the GWLF can simulate large, storm-related pollutant loadings. These loadings may deliver particulate phosphorus that settles to the bottom of the reservoirs and does not affect midlake phosphorus concentration. As originally formulated, the Vollenweider equation does not take into account this fraction of input phosphorus lost to the sediments. Thus, an additional term representing this fraction was added to the Vollenweider model, resulting in Equation 8-6 (NYC DEP, 1999b). where P = phosphorus concentration (mg/L) L = the total P loading to the reservoir from all sources, g/m2/day τw = the mean hydraulic detention time of the reservoir, days H = the mean depth of the reservoir (volume/surface area) fs = the fraction of input phosphorus that is positionally unavailable
OCR for page 321
Watershed Management for Potable Water Supply: Assessing the New York City Strategy Step 2. Determining the contributions of point and nonpoint sources to phosphorus loading has changed dramatically from Phase I. For reservoirs in the Croton watershed, the nested Reckhow model was used to divide the region into multiple subbasins. The nested Reckhow was adopted because NYC DEP felt that the Phase I approach was not accounting for phosphorus retained in nine large waterbodies upstream of the 13 reservoirs. In Phase II, the waterbodies in these subbasins were assumed to retain 50 percent of the phosphorus and were then considered to be point sources to the downstream waterbodies. A 50 percent phosphorus retention rate was used for all of the subbasins in the absence of data on the water budget, phosphorus concentrations, and residence times for these subbasins. Previous studies on reservoirs with residence times greater than six months show that phosphorus retention tends to plateau around 60 percent (K. Kane, NYC DEP, personal communication, 1998). NYC DEP choose 50 percent because although the East-of-Hudson subbasins have longer residence times than the main East-of-Hudson reservoirs, they are also more eutrophic and may be creating phosphorus near the bottom waters as a result of algal decay. As shown in Box 8-2, the Phase II Reckhow model was amended to include some locally measured export coefficients, and data from four years were used rather than data from one year. Another major change for Phase II is that the criterion for determining whether the Reckhow model is a good fit to a basin has been increased from 20 percent to 50 percent in some cases. The 20 percent acceptance criterion was thought to be too restrictive, especially in basins where the phosphorus concentration is very low. In the Catskill/Delaware watershed, the Reckhow model was replaced in favor of the GWLF, which can predict the temporal and spatial variability in phosphorus loading. The GWLF is a numerical model that simulates hydrology, nonpoint source runoff of pollutants, and point source inputs. It is an advancement over the Reckhow approach because it utilizes daily time intervals to generate monthly, seasonal, or annual loadings to reservoirs. Also, it provides separate estimates of groundwater inputs, which are especially important in systems with thin soils such as the Catskill/Delaware reservoirs [in fact, the model was originally developed for the Catskill Region by Haith et al. (1983, 1992) and Haith and Shoemaker (1987) at Cornell University]. Figure 8-1 shows the three submodels that make up the GWLF, and a more detailed description of the model is given in Box 8-3. The model has been calibrated and validated using stream flow data from the Catskill/Delaware reservoirs. Although the data requirements for the GWLF are extensive, literature values are often used for some of the needed parameters. NYC DEP monitoring program provides required data on precipitation, air temperature, land use, soil type, topography, and point source phosphorus loads. There are 11 land use categories in the GWLF as compared to six in the Reckhow model. Data of other model parameters, including soil erodability and pollutant concentrations in ground-
OCR for page 322
Watershed Management for Potable Water Supply: Assessing the New York City Strategy FIGURE 8-1 (A) Water balance submodel of the Generalized Watershed Loading Function. water and surface runoff, are not currently available and have been derived from the literature up to the present time. Step 3. During the Phase I TMDL process, the WLA and LA were calculated only for those basins exceeding their TMDLs. However, during Phase II they were calculated for all basins, regardless of their TMDL status. The methods for calculating the WLA and LA are unchanged from Phase I.
OCR for page 323
Watershed Management for Potable Water Supply: Assessing the New York City Strategy FIGURE 8-1 (B) Dissolved nutrient submodel of the Generalized Watershed Loading Function. Phase III Methodology Although Phase III TMDLs are not due for several years, NYC DEP has been developing more sophisticated models for this process. It is anticipated that the Vollenweider equation will be discarded in favor of a more complex water quality model that has both hydrothermal and eutrophication components. Such a model is currently being designed for the Cannonsville watershed (Cannonsville model), and similar efforts for the other Catskill/Delaware reservoirs are under way. A primary goal of Phase III is to link the GWLF to such a water quality model in order to predict long-term changes in water quality as a result of management practices in the watershed. More information on the hydrothermal and eutrophication models that make up the Cannonsville Model can be found in Auer et al. (1998), Doerr et al. (1998), Owens et al. (1998), and numerous NYC
OCR for page 324
Watershed Management for Potable Water Supply: Assessing the New York City Strategy FIGURE 8-1 (C) Particulate nutrient submodel of the Generalized Watershed Loading Function. DEP publications. As presently formulated, these water quality models focus specifically on nitrogen, phosphorus, dissolved oxygen, and sediment. Water quality and terrestrial models for simulating fate and transport of microbial pathogens and precursors of disinfection byproducts are not currently under development. Phosphorus-Restricted Basins NYC DEP has developed a water quality measure for its reservoirs that is similar in calculation to the TMDL and is used for similar management purposes.
OCR for page 375
Watershed Management for Potable Water Supply: Assessing the New York City Strategy and total trihalomethanes (TTHMs) and the sum of five haloacetic acids (HAA5s) (less than 0.040 mg/L and 0.030 mg/L, respectively). During the season in which the Catskill aqueduct experienced higher turbidity, only DAF/filtration was effective in meeting established water quality goals at the Millford location during the entire year. As a result of Phase I, conventional treatment was eliminated from further consideration because of its high cost and its inability to reliably lower turbidity in the Catskill aqueduct water. Phase II Pilot Plant Study The objective of Phase II of the pilot plant study was to further test and optimize filtration technologies. In addition, consultants were hired to determine how many filtration plants would be necessary and to find a proper location for the treatment plant(s). These activities were conducted simultaneously in order to meet EPA deliverables. First, seven configurations for a treatment facility were designed that included from one to three plants located at several different points along the Catskill and Delaware aqueducts. The water demand was projected to be 1,700 mgd. Second, sites within the watershed region that were suitable for a treatment plant were identified. Of 577 initial locations, most were eliminated by taking into consideration such issues as distance from the aqueducts, acreage of land available, and vacancy of the land. Twelve (12) sites that could not be eliminated from consideration were then combined with the seven configurations to produce 25 treatment schemes. A weighted matrix analysis of the schemes was conducted, taking costs, acceptability, site considerations, flexibility and reliability, implementation, and water quality into account. The preferred treatment scheme consisted of one filtration plant at the Eastview location, which had been identified previously by NYC DEP as a potential site. Eastview is located downstream from Kensico Reservoir and upstream from Hillview Reservoir. Its location would allow water from both the Catskill and Delaware aqueducts to be diverted into the filtration plant. The advantages of the Eastview location are that the appropriate aqueduct connections are possible, the use of Kensico as a settling basin can be maximized, and minimal pumping would be necessary. DAF/filtration and direct filtration were tested further using water derived from sources downstream of Kensico that would be similar to Eastview water. The treatment requirements for the combined ozonation/filtration processes were made more stringent: 99.9 percent inactivation of Cryptosporidium, 99.9 percent inactivation of Giardia, and 99.99 percent inactivation of viruses. It was also required that finished water turbidity be less than 0.10 NTU 95 percent of the time. Ozone doses were chosen that would achieve a 1.5-log removal of oocysts, as determined by a literature review. Filtration was expected to achieve a 1.5-log removal of cysts and oocysts, and this was tested by determining percent removals of surrogate particles of many size ranges during filtration. Overall treatment
OCR for page 376
Watershed Management for Potable Water Supply: Assessing the New York City Strategy was optimized by varying the filtration rate (gallons per minute per square foot), the filter bed depth, the ozonation time and dose, the flocculation time, and the DAF rate (gpm/ft2). Direct filtration achieved the expected 1.5-log removal of particles in the size range of Cryptosporidium (2–30 µm) (Hazen and Sawyer, 1998). The ozone dose was varied between 1.4 and 3 mg/L. FeCl3 was the primary coagulant, a cationic polymer was used (1.5 mg/L), and a filtration rate between 10 and 13 gpm/ft2 was found to be best. In this test, 84 inches of 1.5-mm anthracite were used for filter media. DAF/filtration was also successful in achieving at least a 1.5-log removal of Cryptosporidium-sized (2–30 µm) particles. Because DAF removes natural organic matter prior to ozonation, the required ozone dose was somewhat lower (between 0.75 and 2 mg/L). A filtration rate between 12 and 16 gpm/ft2 and 58 inches of 1.3-mm anthracite for filter media were found to be acceptable. The engineering consultant has recommended, and NYC DEP has approved, one plant at the Eastview location using direct filtration (because of its lower cost) combined with pre-ozonation, as shown in Figure 8-5. The filtration rate would be 13 gpm/ft2 (10 gpm/ft2 in the winter), which would be unprecedented for the East Coast. Hillview Reservoir, the posttreatment storage facility, would need to be covered to protect the superior quality of the filtered water. Cost estimates for construction of the recommended plant range from $2–3 billion (Nickols, 1998) to $4 billion (J. Miele, NYC DEP, personal communication, 1999). A key to this low cost estimate is the use of Hillview as a storage reservoir so that the system does not have to be engineered to meet peak daily demands. Disinfection Study The filtration avoidance determination also includes a study of alternative disinfectants to assess their ability to render inactive Cryptosporidium oocysts in raw water from the Catskill/Delaware system. The motivation for such research was threefold. First, disinfection with chlorine is the primary chemical treatment received by Catskill/Delaware water, but there have been no previous attempts to FIGURE 8-5 Proposed treatment train for direct filtration of the Catskill/Delaware water supply. Source: NYC DEP (1998d).
OCR for page 377
Watershed Management for Potable Water Supply: Assessing the New York City Strategy demonstrate that chlorine is the most effective disinfectant. Second, if regulations for TTHMs and HAA5 in finished water are tightened, then the current reliance on chlorine could take Catskill/Delaware water out of compliance with the Safe Drinking Water Act (SDWA). Third, future regulations for Cryptosporidium log inactivation in water supply systems are likely; a system relying solely on chlorine, in the absence of filtration, would find itself illprepared to comply. The first phase of an ongoing study of alternative disinfectants was recently completed (NYC DEP, 1998e). A major assumption was that the New York City water supply would continue to be of high enough quality to qualify for filtration avoidance. The study evaluated ozone, chlorine dioxide, and chlorine for their disinfecting abilities. However, no actual experiments were performed in which the inactivation of either indicator organisms or pathogens in water samples was measured. Rather, the literature was used to develop target CT (disinfectant concentration multiplied by contact time) values that corresponded to (1) 3-log Giardia inactivation and 4-log virus inactivation (the EPA requirement for unfiltered systems), (2) 1-log inactivation of Cryptosporidium, and (3) 2-log inactivation of Cryptosporidium. These CT values are given in Table 8-4. The suitability of disinfectants was determined solely by comparing the estimated CT values (measured in actual water samples) to the estimated requirements for inactivation of pathogens (gathered from the literature). TABLE 8-4 Target CT Values for the Disinfection Study Developed from Multiple Literature Sources. CT values are in (mg/L)(min). Disinfectant Season (°C) CT 3-log Giardia or 4-log virus 1-log Cryptosporidium 2-log Cryptosporidium Ozone November 1997 (13°) 1.1 6.5 13 March 1997 (7°) 1.9–1.5 8.5 16 July 1998 (19°) 0.8 5 10 Chlorine November 1997 (13°) 86–100 1.4 × 106 March 1997 (5–9°) 139–182 July 1998 (19°) 52–68 3,000–4,000 6,000–8,000 Chlorine November 1997 (13°) 28 Dioxide March 1997 (5°) 33.4 60–80 120–160 July 1998 (20°) 15 30 40 Source: NYC DEP (1998e).
OCR for page 378
Watershed Management for Potable Water Supply: Assessing the New York City Strategy In order to determine achievable CT values for New York City water, decay rates were measured for ozone, chlorine, chlorine dioxide, and disinfectant combinations in water from the Kensico Reservoir Shaft 18 location. Three sets of tests were conducted, using water collected in November 1997 and in March and July 1998. Chlorine decay and chlorine dioxide decay profiles were generated using a variety of initial doses in batch systems. From this, the integrated CT value was computed (using a first-order decay model fit to the data) and was compared to the target levels shown in Table 8-4. Ozone decay experiments were conducted in a continuous-flow sparged pilot plant comprising five columns in series (with ozone application to the first column only). To obtain varying times, the pilot plant was run at different flow rates, and samples were obtained at multiple intermediate times within the ozonation train. Tracer studies were also conducted to obtain the T10 value (time of exit of the fastest 10 percent of the influent). The consultants then used the T10 values at each sampling location along with the observed ozone residual at each point to estimate an apparent "first order" decay rate. The conclusions of the study are that the "CT goals were satisfied for all three conditions established…with reasonable ozone and chlorine dioxide doses and contact times. The applied ozone doses and the contact times used to meet the high CT goal were 1.5 mg/L and 21 minutes in the cold water (5°), and 1.7 mg/L at 25 minutes and 1.9 mg/L at 9 minutes in the warm water (19°). The chlorine dioxide doses and contact times used to meet the high CT goal were 1.0 mg/L at 240 minutes and 1.2 mg/L at approximately 200 minutes in the cold water (5–9°) and 1.7 mg/L at 30 minutes in the warm water (19°). Chlorine disinfection and sequential disinfection (ozone followed by chlorine) were not effective in meeting the established CT goals" (NYC DEP, 1998e). Analysis Use of Literature Review Data. The primary literature used to estimate Cryptosporidium CT values consisted of conference papers by Finch and colleagues, and details of the CT computations were not provided. The use of conference proceedings for CT values is questionable, particularly because there is recent published information on oocyst inactivation. In addition, the limited information on oocyst inactivation used for this study (in which all data were obtained on buffered, demand-free water) is apparently being used to compute a "best estimate" CT value, with no recognition of the uncertainty surrounding this estimate. The CT values computed by EPA ranged from 1.5 to 3 mg/L × min because of varying safety factors. Furthermore, recent work on oocyst inactivation indicates a considerable variability in ozone inactivation efficiency for Cryptosporidium (Oppenheimer et al., 1999), which may result in a large safety factor for regulatory purposes. Hence, the CT values for Cryptosporidium used
OCR for page 379
Watershed Management for Potable Water Supply: Assessing the New York City Strategy in this study may be not on a commensurate scale with the Giardia and viral CT values derived by EPA. Decay Rate Experiments. The methodology used to analyze the ozone decay process is problematic. The use of a semilog plot to extract data from a reactor with backmixing is inappropriate, as it confounds the processes of hydrodynamic dispersion and reaction. Hence, the computed "CT" values achievable via this method are not entirely accurate. A more appropriate methodology to analyze the data should be employed (i.e., use of the full mathematical solution for a reactor with reaction and dispersion). Use of Ozone. The committee questions the use of ozone for treating New York City's drinking water in the absence of other treatment processes. Although ozone is a powerful disinfectant for many microbial pathogens, it can react with dissolved organic matter to produce ketoacids, carboxylic acids, and aldehydes. The reaction of ozone with dissolved organic matter can also change largely refractory humic materials into biodegradable products that can support bacterial regrowth in the distribution system. (It should be noted that regrowth is much less apparent in ozonated systems that carry a distribution system residual provided by another disinfectant.) In addition, in the presence of bromide ion, ozone can lead to formation of bromate and bromine-substituted organic compounds. Because bromide levels were not measured in the disinfection study, the potential for bromate formation in the Catskill/Delaware water supply is currently unknown. To avoid these problems, ozone should only be applied to water with the lowest possible organic content, and ozonation should be followed by a granular media step or secondary disinfection. (The results of using sequential ozone–chlorine disinfection are not encouraging, suggesting that other combinations of disinfectants should be investigated.) Granular media would provide filtration as well as a surface on which heterotrophic organisms can grow and degrade newly formed DBPs and other biodegradable matter. In the absence of these steps, ozonation is not recommended. Conclusions and Recommendations The dual-track approach allows New York City to focus the bulk of its resources on improvements in the watershed. This initial focus will help establish a strong source water protection program without diverting attention and resources toward the details of a filtration plant. The pollution prevention achieved through watershed protection reduces influent pollutant concentrations that would be treated via filtration. If the source water protection program is effective, the cost of filtration can be reduced.
OCR for page 380
Watershed Management for Potable Water Supply: Assessing the New York City Strategy The results of the filtration pilot plant study show that the present New York City water supply can be effectively treated by ozonation combined with coagulation/filtration. Treated water from direct filtration had a turbidity of less than 0.1 NTU, average particle (2–30 µm) counts ranging from 8 to 153/mL, and total trihalomethane and total trihalomethane and haloacetic acid formation potential of less than 0.040 mg/L and 0.030 mg/L, respectively. At least 3-log oocyst inactivation/removal is expected for the entire treatment train. These low effluent pollutant concentrations from a potential filtration plant and dependent on maintaining high source water quality via aggressive watershed management. The construction cost of such a treatment facility ranges from $265 to $400 per capita. New York City should conduct studies on the actual inactivation of pathogens in its water under potential design conditions. In view of the potential effect of as yet unknown water quality factors on inactivation efficiency, and in view of the large potential investment that enhanced disinfection might require, it is not prudent to rely upon literature values for oocyst inactivation efficiency. These studies should be conducted using best available methodology for assessing cyst, oocyst, and virus viability and for virus viability and for susceptibility testing Additional studies to assess the potential of ozone as a treatment technique are required. Any consideration of ozonation should include measurements of bromide in the source waters to determine the potential for bromate formation. The literature to date suggests that ozone has the potential to increase biodegradable organic carbon in finished water and to foster regrowth of heterotrophic plate count organisms and possibly coliforms, although distribution system disinfectant residuals may counter this phenomenon. Without assurance that such regrowth would occur, it is imprudent to consider ozone as a sole treatment method. A decision to construct a filtration plant should in no way deter New York City from pursuing an aggressive watershed management program. If a coagulation/filtration plant is put in place, it should be treating the best-quality source water possible. For that reason, high water quality in the Catskill/Delaware system must be maintained via aggressive implementation of the watershed management strategy. REFERENCES Anderson, W.P., J.A. Klang, and R. Peplin. 1997. Water pollutant trading: From policy to reality. Environmental Engineer October. Auer, M.T., S.T. Bagley, D.A. Stern, and M.J. Babiera. 1998. A framework for modeling the fate and transport of Giardia and Cryptosporidium in surface waters. Journal of Lake and Reservoir Management 14(2-3):393–400.
OCR for page 381
Watershed Management for Potable Water Supply: Assessing the New York City Strategy Azous, A. L., L. E. Reinelt, and J. Burkey. 1997. Management of freshwater wetlands in central Puget Sound Basin. Chapter 12 In Zous, A. L., and R. R. Horner (eds.) Wetlands and Urbanization: Implications for the Future. King County, WA: Department of Natural Resources Water and Land Resources Division. Brown, W., and T. Schueler. 1997. Pollutant Removal of Urban Stormwater Best Management Practices—A National Database. Ellicott City, MD: Center for Watershed Management, Chesapeake Research Consortium. Caraco, D., J. Zielinski, and R. Claytor. 1998. Nutrient Loading from Conventional and Innovative Site Development. Ellicott City, MD: Cheseapeake Research Consortium, Center for Watershed Protection. Claytor. 1996. Stormwater Retrofits–A tool for watershed management. Technical Note 48 Watershed Protection Techniques 1(4):188–192. Clean Water Report. 1998. Nation's first pollutant trading water permit approved by EPA Region I. Clean Water Report 36(23):224. Cronin, R. 1997. NYS DEC. Presentation given at the first meeting of the NRC Committee, September 25–26, 1997, New York City, NY. Dillon, P. J., and F. H. Rigler. 1974. Limnology and Oceanography 19:767. Doerr, S. M., E. M Owens, R. K. Gelda, M. T. Aurer, and S. W. Effler. 1998. Development and testing of a nutrient-phytoplankton model for Cannonsville Reservoir. Journal of Lake and Reservoir Management 14(2-3):301–321. Environmental Protection Agency (EPA). 1988a. Memorandum on State Antidegradation Program Review to R. L. Caspe, Region II Director of Water Management Division from M. G. Portho, Director of the Office of Water Regulations and Standards. September 12. EPA. 1988b. Memorandum on EPA's Review of New York's Toxic Control Program to D. Barolo, Director of Division of Water, NYS DEC from R. L. Caspe, Region II Director of Water Management Division. September 30. EPA. 1993a. Subsurface Flow Constructed Wetlands for Wastewater Treatment: Design Manual Number 74. EPA 832-R-93-008. Washington, DC: Office of Water. EPA. 1993b. Memorandum on State Antidegradation Implementation Methods to Regional Water Quality Standards Coordinator from B. Shippen, Water Quality Standards Branch. May 26. EPA. 1994. Water Quality Standards Handbook. Washington, DC: EPA. EPA. 1996. Draft Framework for Watershed-based Trading. EPA 800-R-96-001. Office of Water. Washington, DC: EPA. EPA. 1997. Compendium of Tools for Watershed Assessment and TMDL Development. EPA-841-B-97-006. Office of Water. Washington, DC: EPA. Gain, W. 1996. The Effects of Low Flow Path Modification on Water Quality Constituent Retention in an Urban Stormwater Pond and Wetland System. USGS Water Resources Investigations Report 95-4297. Orlando, FL: USGS. Haith, D. A., and L. L. Shoemaker. 1987. Generalized watershed loading functions for stream flow nutrients. Water Resources Bulletin 23(3):471–478. Haith, D. A., and L. L. Shoemaker, R. L. Doneker, and L. D. Delwiche. 1983. Non-point Source Control of Phosphorus–A Watershed Evaluation, Volume 4, Modeling of Streamflow, Sediment Yield, and Nutrient Export from the West Branch Delaware River Watershed, U.S. Environmental Protection Agency, ORD, EPA-P002-137-02. Haith, D. A., R. Mandel, and R. S. Wu. 1992. GWLF Generalized Watershed Loading Functions, Version 2.0 User's Manual. Ithaca, NY: Department of Agricultural & Biological Engineering, Cornell University. Harding, D. 1993. TMDL Case Study: Tar-Pamlico Basin, North Carolina, Washington, DC: EPA. Hazen and Sawyer. 1996. Phase I Pilot Testing Program Preliminary Summary of Results. June 18, 1996.
OCR for page 382
Watershed Management for Potable Water Supply: Assessing the New York City Strategy Hazen and Sawyer. 1998. Task 4: Phase II Pilot Testing Report. EPA Submittal. Corona, NY: NYC DEP. Hoag, D. L., and J. S. Hughes-Popp. 1997. The Theory and Practice of Pollution Credit Trading in Water Quality Management. Review of Agricultural Economics, 19(Fall/Winter): 252–262. Izeman, M. A. 1998. NRDC. Memorandum to the National Research Council dated May 13, 1998. Izeman, M. A., and R. Marx. 1996a. Comments of the Natural Resources Defense Council, Inc. on the New York State Department of Environmental Conservation's Proposed Phase I Total Maximum Daily Loads for Phosphorus in the Croton River System. November 29, 1996. Izeman, M. A., and R. Marx. 1996b. Comments of the Natural Resources Defense Council, Inc. on the New York State Department of Environmental Conservation's Proposed Phase I Total Maximum Daily Loads for Phosphorus in the Catskill and Delaware Systems. December 27, 1996. Izeman, M. A., and R. Marx. 1998. Comments of the Natural Resources Defense Council, Inc. on the New York City Department of Environmental Protection's Draft Methodology for Calculating Phase II Total Maximum Daily Loads (TMDLs) of Phosphorus for New York City Drinking Water Reservoirs. Jacobson, E. M., D. L. Hoag, and L. E. Danielson. 1994. Theory and Practice of Pollution Credit Trading in Water Quality Management. The Colorado Agricultural Experiment Station and the North Carolina Agricultural Research Service. Janus, L. 1989. The relationship of nutrient residence times to the trophic condition of lakes. Ph.D. dissertation, McMaster University, Ontario. Kadlec, R. H., and R. L. Knight. 1996. Treatment Wetlands. Boca Raton, FL: Lewis Publishers. Knight, R. L., R. Randall, M. Girst, J. A. Tess, M. Wilhelm, and R. H. Kadlec. 1995. Arizona Guidance Manual for Constructed Wetlands for Water Quality Improvement. Arizona Department of Environmental Quality. Krudner, M. 1997. EPA Region 2. Presentation given at the first meeting of the NRC Committee, September 25-26, 1997, New York City, NY. Lapp, D. 1994. Managing market mania. Environmental Action Winter:17-21. Larsen, D. P., J. VanSickle, K. W. Maleug, and P. D. Smith. 1976. Water Research 13:1259–1268. Longabucco, P., and M. R. Rafferty. 1998. Analysis of Material Loading to Cannonsville Reservoir: Advantages of Event-Based Sampling. Journal of Lake and Reservoir Management 14(2-3): 197–212. Mann, E. 1994. Trading delusions. Environmental Action Winter:22–24. Maryland Department of Environment (MDE). 1998. Stormwater Design Manual. Volume 1 and 2. Baltimore, MD: Water Management Administration. McCann, K., and L. Olson. 1994. Pollutant removal efficiencies of the Greenwood Urban Wetland Treatment System. Journal of Lake and Reservoir Management 9(2):97–106. New York City Department of Environmental Protection (NYC DEP). 1993. Implications of Phosphorus Loading for Water Quality in NYC Reservoirs. Corona, NY: NYC DEP. NYC DEP. 1996. Methodology for Calculating Phase II TMDLs of Phosphorus for NYC Drinking Water Reservoirs. December 31, 1996. Valhalla, NY: NYC DEP. NYC DEP. 1997a. Methodology for Determining Phosphorus Restricted Basins. July 1997. Valhalla, NY: NYC DEP. NYC DEP. 1997b. The Relationship between Phosphorus Loading, THM Precursors, and the Current 20 µg/L TP Guidance Value. Valhalla, NY: NYC DEP. NYC DEP. 1997c. Guidance for Phosphorus Offset Pilot Programs. April 1997. Valhalla, NY: NYC DEP. NYC DEP. 1997d. Guidance for Preparing a Variance Application for a WWTP Expansion Involving a Phosphorus Offset Plan. May 1997. Valhalla, NY: NYC DEP. NYC DEP. 1998a. Calibrate and verify GWLF models for remaining Catskill/Delaware Reservoirs. January 27, 1998. Valhalla, NY: NYC DEP.
OCR for page 383
Watershed Management for Potable Water Supply: Assessing the New York City Strategy NYC DEP. 1998b. NYC DEP Review of the Phase I TMDL Implementation. June 29, 1998. Valhalla, NY: NYC DEP. NYC DEP. 1998c. Report on Applications to Participate in the Pilot Phosphorus Offset Program. January 1998. Valhalla, NY: NYC DEP. NYC DEP. 1998d. Capital Project No. WM-30 Catskill and Delaware Water Treatment: Treatment Process Selection to Commence Preliminary Design. NYC DEP. 1998e. Disinfection Study for the Catskill and Delaware Water Supply Systems. October 30, 1998. Valhalla, NY: NYC DEP. NYC DEP. 1999a. Development of a water quality guidance value for Phase II Total Maximum Daily Loads (TMDLs) in the New York City Reservoirs. Valhalla, NY: NYC DEP. NYC DEP. 1999b. Methodology for Calculating Phase II Total Maximum Daily Loads (TMDLs) of Phosphorus for New York City Drinking Water Reservoirs. March 1999. Valhalla, NY: NYC DEP. NYC DEP. 1999c. Proposed Phase II Phosphorus TMDL Calculations for Ashokan Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999d. Proposed Phase II Phosphorus TMDL Calculations for Cannonsville Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999e. Proposed Phase II Phosphorus TMDL Calculations for Neversink Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999f. Proposed Phase II Phosphorus TMDL Calculations for Pepacton Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999g. Proposed Phase II Phosphorus TMDL Calculations for Rondout Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999h. Proposed Phase II Phosphorus TMDL Calculations for Schoharie Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999i. Proposed Phase II Phosphorus TMDL Calculations for Kensico Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999j. Proposed Phase II Phosphorus TMDL Calculations for West Branch Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999k. Proposed Phase II Phosphorus TMDL Calculations for Amawalk Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999l. Proposed Phase II Phosphorus TMDL Calculations for Bog Brook Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999m. Proposed Phase II Phosphorus TMDL Calculations for Boyds Corner Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999n. Proposed Phase II Phosphorus TMDL Calculations for Cross River Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999o. Proposed Phase II Phosphorus TMDL Calculations for Croton Falls Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999p. Proposed Phase II Phosphorus TMDL Calculations for Diverting Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999q. Proposed Phase II Phosphorus TMDL Calculations for East Branch Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999r. Proposed Phase II Phosphorus TMDL Calculations for Middle Branch Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999s. Proposed Phase II Phosphorus TMDL Calculations for Muscoot Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999t. Proposed Phase II Phosphorus TMDL Calculations for New Croton Reservoir. Valhalla, NY: NYC DEP. NYC DEP. 1999u. Proposed Phase II Phosphorus TMDL Calculations for Titicus Reservoir. Valhalla, NY: NYC DEP.
OCR for page 384
Watershed Management for Potable Water Supply: Assessing the New York City Strategy New York State Department of Environmental Conservation (NYS DEC). 1985. Water Quality Antidegradation Policy. Albany, N.Y.: Division of Water. NYS DEC. 1990. Director Pagano on Antidegradation. NYSDEC Water Bulletin August 1990. NYS DEC. 1991. Final Combined Supplemental Environmental and Regulatory Impact Statement for Revision of Water Quality Regulations: Surface Waters and Groundwaters. 6 NYCRR Parts 700-705. NYS DEC. 1992. SEQR Handbook. Albany, N.Y.: Division of Regulatory Affairs. NYS DEC. 1993a. New York State Fact Sheet: Ambient Water Quality Value for Protection of Recreational Uses. Albany, N.Y.: Bureau of Technical Services and Research. NYS DEC. 1993b. Proposed Antidegradation Review Process for New York State: Community Assessment Form. Albany, N.Y.: Division of Water. NYS DEC. 1996. Urban/Stormwater Runoff Management Practices Catalogue for Nonpoint Source Pollution Prevention and Water Quality Protection in New York State. November, 1996. NYS DEC. 1997. Implementation of the NYSDEC Antidegradation Policy—Great Lakes Basin (Supplement to Antidegradation Policy dated September 9, 1985). Division of Water Technical and Operational Guidance Series 1.3.9. June 30. NYS DEC. 1998. Total Maximum Daily Loads and Water Quality—Based Effluent Limits. Division of Water Technical and Operations Guidance Series 1.3.1. February 26. Nickols, D. 1998. Hazen and Sawyer. Presentation given at the Fourth Meeting of the NRC Committee. August 31–September 1, 1998, Washington, DC. Novotny, V. 1996. Evaluation and Comments on Proposed Phase I TMDL Calculations for Catskill-Delaware Reservoir. Mequon, W.I.: Aqua Nova International, Ltd. Oberts, H. 1997. Lake McCarrons Wetland Treatment System: Phase III Study Report. St. Paul, MN: Metropolitan Council Environmental Services. Oppenheimer, J. A., E. M. Aieta, R. R. Trussell, J. G. Jacangelo, and I. N. Najm. 1999. Evaluation of Cryptosporidium Inactivation in Natural Waters. Denver, CO: AWWA Research Foundation. Draft. Owens, E. M. 1998. Development and testing of one-dimensional hydrothermal models of Cannonsville Reservoir. Journal of Lake and Reservoir Management 14(2-3):172–185. Paulson, C. L. 1997. Testimony on Cherry Creek Basin Water Quality Authority before the Water Quality Control Commission of the State of Colorado. Reckhow, K. H., M. N. Beaulac, and J. T. Simpson. 1980. Modeling phosphorus loading and lake response under uncertainty: A manual and compilation of export coefficients. Vol. 440/5-80-011. Washington, DC: EPA. Reckhow, K. H., and Chapra, S.C. 1983. Engineering Approaches for Lake Management, Vol. 1—Data Analysis and Empirical Modeling, Boston, MA: Butterworth Publishers, Reed, S. C., E. J. Middlebrooks, and R. W. Crites. 1995. Natural Systems for Waste Management and Treatment, McGraw Hill, 2nd Edition. Richter, K. O., and A. L. Azous. 1995. Amphibian occurrence and wetland characteristics in the Puget Sound Basin. Wetlands 15(3):305–312. Sawyer, C. N. 1947. J. New England Water Works Association 61:109. Schnoor, J. L. 1996. Environmental Modeling—Fate and Transport of Pollutants in Water, Air and Soil. New York, NY: John Wiley and Sons. Schueler, T. 1987. Controlling Urban Runoff—A Manual for Planning and Designing Urban Best Management. Washington, DC: Metropolitan Washington Council of Governments. Schueler, T. 1996. Irreducible pollutant concentrations discharged from urban best management practices. Watershed Protection Techniques 2(2):369–372. Stepczuk, C. L., A. B. Martin, P. Longabucco, J. A. Bloomfield, and S. W. Effler. 1998a. Allochthonous contributions of THM precursors in a eutrophic reservoir. Journal of Lake and Reservoir Management 14(2-3):344–355.
OCR for page 385
Watershed Management for Potable Water Supply: Assessing the New York City Strategy Stepczuk, C. L., A. B. Martin, S. W. Effler, J. A. Bloomfield, and M. T. Auer. 1998b. Spatial and temporal patterns of THM precursors in a eutrophic reservoir. Journal of Lake and Reservoir Management 14(2–3): 356–366. Stepczuk, C. L., E. M. Owens, S. W. Effler, J. A. Bloomfield, and M. T. Auer. 1998c. A modeling analysis of THM precursors for a eutrophic reservoir. Journal of Lake and Reservoir Management 14(2-3): 367–378. Stephenson, K. 1994. Effluent Allowance Trading: A New Approach to Watershed Management. Blacksburg, VA: Virginia Water Resources Research Center. The Cadmus Group, Inc. 1998. Technical Analysis of State and Regional Antidegradation Procedures. Final Draft. Prepared for the U.S. Environmental Protection Agency under Contract 68-C4-0051. Waltham, MA: The Cadmus Group, Inc. Urbonas, B. 1994. Performance of a pond/wetland system in Colorado. Technical Note 17. Watershed Protection Techniques 1(2):68–71. Vollenweider, R. A. 1976. Advances in defining critical loading levels for phosphorus in lake eutrophication. Mem. Ist. Ital. Idrobiol. 33:53–83. Warne, D. 1999. NYC DEP. Memorandum to the National Research Council dated April 1999.
Representative terms from entire chapter: