 |
|
|
|||||||||||||||||||||||||||||||||||||||
 
|
Part 2: Design criteria common to all filter types
Part 1 of this article in the March/April issue of Stormwater covered terms and terminology. As noted in Part 1, I have observed in manuals, articles, and reports, as well as in presentations and conversations at conferences, that the complexity of terminology itself leads to misperceptions and confusion over expected performance, as well as to unnecessary and inappropriate distinctions in design procedures and criteria. This dynamic has led to inconsistencies in design procedures, frequently within the same manual. This may result in a bias toward particular systems, because design criteria drive costs. It also complicates the design process for practitioners, particularly for those in local government who review the drainage plans of the development engineer. Presented in Part 1 of the article was Table 1, summarizing the many names and the widely varying design criteria. I suggested simplification. Two scenarios were offered, repeated in Part 2 as Tables 2 and 3. Engineers do not necessarily realize these differences and potential conflicts because we work within our own communities, states, or provinces with agreed terminologies and sets of design criteria. However, as it becomes increasingly common to trade experiences and field results across regions and borders, contradictions and miscommunication are becoming more frequent. A common and simplified set of terminology and design procedures is warranted. Certainly the design procedures should be consistent within a given manual. Parts 2 and 3 of this series cover design criteria. Part 2 covers design criteria common to all filter types. Part 3 covers design criteria specific to each filter type. It is not my intent to suggest that all manuals should use the same design criteria. Rather, it is that design criteria should be consistent throughout any given manual. We can expect differences between manuals because of differing judgments of manual authors. There are, of course, regional differences to consider when selecting design criteria but most particularly climate as noted in Parts 2 and 3 Design Criteria
Regardless of the filter type:
Varying with filter type:
The first set of criteria is covered in Part 2. The second set will be covered in Part 3. Surface Vegetation What are the benefits of vegetation? Aesthetics, of course. It is commonly believed that plants are significant removers of pollutants through growth. However, their contribution, while significant, is primarily indirect. Plant roots help retain if not enhance filtration or infiltration rates, particularly in tighter soils. These observations are reinforced by agricultural field studies. Deep-rooted prairie grass has been found to have a greater effect than shallow-rooted turf (Bannerman 2006). Vegetation increases water-quantity reduction by evapotranspiration, in turn reducing the temperature of the remaining discharge, of benefit to cold-temperature streams. The relative roles of infiltration versus evapotranspiration need further definition. A caution with vegetation but particularly turf grass is overwatering and excessive use of fertilizers. These may occur where aesthetics is of particular concern. Plants capture incoming sediment that would otherwise clog the filter surface. Anecdotal observations indicate turf grass significantly extends the period before maintenance is required of sand filters and likely does so for bioretention filters and infiltration basins (Minton 2005). It is easier and likely less costly for homeowner associations to mow or prune than to remove sediment. Through their death and decay, plants contribute organic matter that sorbs dissolved metals, pesticides, and other toxic organics. The organic matter also serves as food substrate for bacteria important to the degradation of pesticides, petroleum, polycyclic aromatic hydrocarbons, pathogenic bacteria and viruses, and nitrogen. Degradation is more rapid as the vegetation promotes a diverse ecosystem of degraders. Vegetation might minimize the tendency of media to freeze, particularly by preventing the formation of the thin sediment layer as occurs with bare infiltration and sand filters. Recommendation: Vegetation should be required of all surface filter types where biologically feasible. Use native shrubs and trees and minimize the use of turf grass given its maintenance costs and the tendency of owners to fertilize and water excessively. Limited data suggest that deep-rooted plants like prairie grasses provide an incremental benefit over the shallow-rooted turf grasses for infiltration in tighter soils. It is not clear that this is the case with coarse media as in bioretention filters. Deep-rooted plants likely enhance evapotranspiration rates in these filter types. Whether these effects are sufficient to warrant distinctive names for shallow- and deep-rooted plant cover is not yet clear. There are limitations to vegetative covers in semiarid regions. However, I have observed vegetated filters and grass swales in southern California and Boise, ID. Vegetation was maintained by excess water entering the facilities from overwatered lawns. Even partial coverage with shrubs is of benefit according to one field study. Sizing the Filter Surface Area
Darcy’s Law is a hydraulic equation: It gives the filter area needed to pass the DWQV within the specified drawdown time. Good performance is presumed by the specification of media type, gradation, and thickness. Darcy’s Law includes the effect of the thickness or depth of the media. Although the parameter appears in both the denominator and the numerator, the thicker the media, the greater the surface area required. However, the layer of accumulated sediment on the surface of the filter controls the flow rate, not the media itself, and therefore media thickness is likely irrelevant (Minton 2005). Alternative equations that recognize this point have been proposed (Urbonas 1999, Clark and Pitt 1999), but more studies are needed, particularly at different operating water depths. The DWQV is commonly determined by multiplying a specified rainfall depth times the drainage area times a runoff coefficient. The specified runoff depth differs between manuals due to climatic differences and differences in the volume performance goal. Some manuals specify a volume goal: for example, that at least 90% of the runoff generated over time be treated. Others specify that a percentile storm be treated: for example, a storm depth representing the 90th percentile—that is, the rainfall depth represents 90% of the events. A few manuals determine the area of bioretention cells by the simple expedient of either dividing the DWQV by the specified maximum depth or specifying a minimum area as a percentage of the drainage area, usually 2% to 4%. One manual specifies a minimum of 20%. Darcy’s Law is not used to determine the area of a dry swale. Like bioretention (in some manuals), its area is determined by dividing the DWQV by the specified average operating water depth of 12 inches. Recommendation: Use Darcy’s Law to size all filter types listed in Table 4, including infiltration basins and trenches. Equation 1 presumes that the entire DWQV must be stored within the basin. The optimal filter area and system volume simplification is not likely achieved with this simplification. Instead, we should view the live volume in the system, both the filter and the pretreatment unit, as an equalization basin that evens out the significantly varying inflow rate. It need not necessarily be equal to the WQV. Other factors should be considered, including variable inflow rates (not within but from storm to storm), the varying interval periods between storms, evaporation and evapotranspiration where vegetation is present, and that infiltration/filtration occurs during each storm. Particularly for wet climates, it is preferable to use continuous simulation, incorporating the actual historic rainfall record to generate runoff sequences (WDOE 2005) rather than the simple runoff depth as is commonly used. With continuous simulation, the practitioner can explicitly incorporate the key variable factors mentioned above. Mounding can and should also be explicitly considered with continuous simulation relevant to infiltration systems. A complementary method to determine the surface area is to consider directly the effect of accumulated sediment on the reduction of the hydraulic conductivity (Urbonas 1999, Minton 2005). The procedure has two steps: calculating the annual accumulation of sediment on the filter and then calculating the surface area. The two steps are presented as Equations 2 and 3, respectively. Equation 3 directly incorporates consideration for maintenance.
Equation 2 is a simple mass balance of sediment accumulation based on influent concentration. With simple modification, Equation 2 can be based on area loading as pounds per acre per year. Ideally, the engineer wants to know the accumulation specific to the design hydraulic conductivity, the point in time when the filter should be cleaned. Limited data suggest 0.25 to 1 pound per square foot (Urbonas 1999, Minton 2005), too broad at this time for practical use. There is not likely one value applicable to all geographic regions due to the effects of differing sediment characteristics and climate. As the percentage of clay increases, the allowable accumulation likely decreases. The allowable accumulation is likely greater in semiarid regions where extended periods between storms desiccate bacteria and algae growth and dry the sediment. The results from Equations 1 and 3 are compared; the larger surface area is used for design. Hydraulic Conductivity and Filtration or Infiltration Rate Current design values for Hc vary among manuals from unspecified to 0.5 foot per day for bioretention filters, 2 to 4 feet per day for sand filters, and 8.7 feet per day for organic or peat filters. Very few manuals specify 20 feet per day for sand filters. The Hc is infrequently specified for infiltration systems. Clean sand has a Hc of about 2 to 5 feet per hour (not per day). The design criterion of 2 or 4 feet per day is substantially lower. Why? The specified Hc represents a relatively clogged filter, the condition when maintenance should occur. In effect, the values commonly selected for sand filters consider maintenance. Hence, those few manuals that specify 20 feet per day are not considering the effect of sediment accumulation. It is seen with Equation 1 that the very high Hc value reduces the filter area by a factor of 5 to 10, with respect to other manuals. Smaller filters will clog sooner, requiring more frequent maintenance and/or a failure to meet the stated volume performance goal. Filtration/infiltration rate is the rate of movement of water down through the media. It is not constant, decreasing as the water level drops above the filter surface. Filtration/infiltration rate is often confused with hydraulic conductivity, perhaps because they have the same units. However, Hc is not affected by the water level. Using Darcy’s Law, the Hc and filtration/infiltration rate are related for a unit area of filter by Equation 4.
Filtration or infiltration rates commonly specified are 1 inch to 3 inches per hour for a bioretention cell and 0.5 inch per hour minimum for dry swales and infiltration systems. A few manuals specify a maximum infiltration rate, ranging from 2 to 9 inches per hour. In these cases, full treatment is required prior to entry into the infiltration unit. The reasoning is that the high infiltration rate implies a coarse soil lacking organic matter, raising concern about the treatment effectiveness of the soil. Infiltration rate and hydraulic conductivity are essentially the same when the hydraulic gradient is about 1, which occurs with bioretention filters with a shallow operating water depth relative to the depth of the media, and infiltration basins where the separation to groundwater is at least 10. Confusion between hydraulic conductivity and filtration/infiltration rate has led to inconsistencies in manuals. These inconsistencies include specifying different values for hydraulic conductivity for bare surface filters with different media; not recognizing that it is the accumulated sediment that eventually controls flow rate, rather than the media; specifying hydraulic conductivity for bioretention filters but infiltration rate for dry swales, although a swale is essentially a sloped bioretention filter; specifying values for both parameters, not recognizing their relationship, with the potential for conflict; and recognizing gradual clogging in the sizing of sand filters but not of infiltration basins. About a quarter of the manuals reviewed specify an adjustment factor to the infiltration rate to recognize the likely variability of the measured infiltration rate across the proposed basin. But only one manual factors into sizing the expectation of gradual clogging (WDOE 2005). This results in the oddity that there is an adjustment factor for sand filters that is effectively about 20 but no adjustment factor when sizing infiltration basins in sandy soils. Turf grass in sod placed on sand reduces the hydraulic conductivity relative to sand but not likely below that of the effect of the accumulated sediment layer. Vegetation appears to maintain the hydraulic conductivity of sand and bioretention filters and likely improves it for infiltration basins in relatively tight soils. More field data are needed. Recommendation: Use hydraulic conductivity with Darcy’ Law to size all filter types, and use a consistent value. Current values for sand filters of 2 to 4 feet per day are reasonable given our limited understanding of the effect of accumulated sediment. It is also reasonable to use the same specification for bioretention filters with the sand/organic blend and for infiltration basins placed in sandy and coarser soils. That is, if the manual author concludes that 4 feet per day is a reasonable design specification, it can be used for all filter types given that it is the accumulated sediment that controls flow and not the underlying media. With respect to infiltration basins, intermediate values between 0.5 and 2 (to 4) would be used for the intermediate soils having infiltration rates between 0.5 inch and 2 inches per hour, where we are in the transition area as to whether flow rate is controlled by the accumulated sediment or the native soil. These design coefficients are for saturated conditions in the media or the soil. Saturation may not occur with tighter soils, such as loamy sand beneath infiltration basins. It is proposed the commonly used minimum infiltration rate of 0.5 inch per hour continue to be used to limit the placement of infiltration systems but that Darcy’s Law be used to size the surface area. There are likely climatic considerations. Hydraulic conductivity is dependent on water viscosity, which decreases with decreasing temperature and increasing chloride content, of interest in cold-climate regions. The viscosity of water at 5oC is about 70% greater than at 20oC; hence, the Hc is 70% lower. We have not been taking temperature into consideration in establishing our criteria. The most commonly specified criterion of 3.5 feet per day was developed in Austin, TX. One study found the Hc during the winter to be half that observed during the summer (Heasom, Traver, and Welker 2006).
It is likely that wet (temperate and humid) regions should use a lower value than semiarid areas because of the effect of prolonged wetness of the filter media. Uncertain is what the value should be with turf grass as the cover. Should it be the same as for loam soil or within the range noted above, as it represents an essentially clogged filter? We need to better define the relationship of the hydraulic conductivity, presence or absence of vegetation, water temperature, sediment accumulation, and prolonged wetness. We also need to better understand the effect of the characteristics of the accumulated sediment: percentage of clay versus silts and sand, organic matter, and biological mats. Operating Water Volume Most manuals are inconsistent in specifying the water volume between filter types. But the nature of the inconsistency varies between manuals. Essentially all manuals specify that an infiltration basin or a trench be sized to retain the WQDV. For sand filters, most manuals are unclear and do not appear to recognize that the volume above the filter area, and pretreatment unit if specified, may be less than the WQDV. A few manuals specify that the operating volume for sand filters be equal to 75% of the WQDV, apparently recognizing that water is moving through the filter during the storm, but do not give the same recognition to the other filter types. In contrast, some manuals require that the sand filter retain the DWQV but not apparently bioretention filters, given the sizing procedure. For bioretention, these manuals specify a water depth and filter area as a percentage of the drainage area, which results in a storage volume that is less than the DWQV. Some manuals include the open volume in the media in the calculation for bioretention filters but not for sand filters or dry swales. As the three filter types usually have similar filtration capacities, the specification as to storage should be consistent within a manual. These adjustments are not unreasonable if we wish to recognize that stormwater is draining through the facility during the storm. However, this may not be prudent in wet climates: Water may still be in the basin when the next storm arrives. Recommendation: Barring the more definitive analysis of continuous simulation as previously suggested, it is advisable to specify the operating volume be equal to the DWQV for all filter types. This recommendation is certainly advisable for wet climates. A lesser volume is reasonable in semiarid climates with extended periods between storms but should be consistent between filter types. It has been noted that Darcy’s Law generally produces a filter area that is smaller than the area needed to store the DWQV. The operating volume drives the total facility footprint. This is commonly the case for sand filters but may also be the outcome for bioretention units depending on the specified drawdown time and operating depth. In such cases, the bioretention unit need cover only a portion of the total facility as with sand filters. In such cases, the vegetated storage area provides the pretreatment function. Maximum Drainage Area Recommendation: It is recommended that no limit be placed on the drainage area. If a manual author prefers to do so, the limit should be the same for all filter types. Aesthetics and the particular application will result in practical limits for each filter type—for example, the lineal sand filter and the bioretention filter. Summary References Gary Minton, Ph.D., P.E., is an independent consultant on stormwater treatment with Resource Planning Associates in Seattle. He is author of the book Stormwater Treatment: Biological, Chemical, and Engineering Principles. SW May 2008 |
||||||||||||||||||||||||||||||||||||||||
 |
|
|
Home + About + Subscribe + News + Calendar + Glossary Distributed Energy | Erosion Control | Grading & Excavation Contractor | MSW Management | Onsite Water Treatment | Water Efficiency | StormCon | ForesterPress | Forester Media
|
