The Utah Department of Transportation (UDOT) conducted a study to work with the UDOT Technical Advisory Committee (TAC) to prepare a selection methodology and performance-based specification for hydrodynamic separators and oil-water separators as a structural control measure for storm water treatment. The following article is a summary of the UDOT Oil Water Separator Study; the full report is available on the department’s website.
The study investigated the following elements for the design and implementation of water-quality BMPs:
- Identify the storm water pollutants of concern;
- Define the characteristics of sediment in storm water runoff;
- Evaluate the effectiveness of the control to remove hydrocarbons; and
- Prepare a selection methodology for water-quality BMPs.
Recognizing the difference in storm water quality given the different land uses is beneficial to developing a storm water management plan and implementing proper BMPs. Generally, storm water tends to contain similar pollutants; however, different land uses will contribute different amounts of each contaminant.
Storm water discharges from construction sites were targeted as a major source of sediment to receiving waters. Sediment, metals, oil and grease concentrations are increased in storm water due to impervious surfaces and the presence of automobiles.
Some studies have investigated storm water quality with relation to average annual daily traffic (AADT). Storm water runoff from residential, commercial and industrial park land uses can be generally characterized. These land uses represent a majority of the area in urban centers. The NURP database, as previously stated, indicates that the most prevalent pollutants from these land uses are solids, metals, floatables (litter) and nutrients.
Characteristics of sediment
Suspended solids are one of the most common contaminants found in urban storm water. Available studies indicate that most particles suspended in storm water are less than 120 µm in diam. Coarser fractions, above 120 µm, tend to remain in gutters or get caught in catchbasins. Sediment coarser than medium silt (~20 µm) settles rapidly, but much longer detention times are required for finer particles to settle. Particles less than 10 µm tend not to settle discretely, according to Stokes Law, and flocculants are required to aid in particle settling. The particle shape, density, water viscosity, electrostatic forces and flow characteristics affect settling rates. Figure 1 (S22) provides information regarding particle size distribution and laboratory settling rates from a study in New Zealand.
Typical hydrocarbon concentrations
In considering use of OWSs for treating storm water, it is important to evaluate documented oil/grease or hydrocarbon concentrations in urban storm water runoff. Current literature and/or studies indicate that oil and grease concentrations in urban storm water runoff are generally low. Table 1 (S22) presents the information obtained during a literature search.
In a study designed to evaluate the performance of separators, an analysis of droplet size was provided. Oil/water mixtures are usually divided into four categories:
- Free-floating oil (droplet sizes of 250 µm or more), oil slick or film;
- Oil droplets and globules ranging in size from 10 to 300 µm;
- Emulsions (1 to 30 µm range) - Stable: usually the result of surfactants holding droplets in solution; and
- Non-stable: created by shearing forces present in mixing; and
- Dissolved oil (<10 µm) *(µm=micron)
The objective of OWSs is to treat most of the flow (90 to 95%) from the catchment to an acceptable degree (10 to 15 mg/L O&G) and to remove fee oil so as not to produce a discharge that causes an ongoing or recurring visible sheen. These systems are not effective at removing emulsified or dissolved oils. Oil droplets are generally characterized as emulsified or dissolved (1 to 30 µm range) in municipal storm water discharges (Table 1).
At this time, literature that suggests typical oil droplet sizes or size distribution found in storm water runoff has not been found. Hydrodynamic separator vendors indicate in sales and technical literature (some based on data from petroleum storage terminals) that 80% of the droplets (by volume) are greater than 90 mm, and that 30% are greater than 150 µm.
Typically, effluent oil and grease concentrations from separators can meet 10 to 20 mg/L, which generally corresponds to the removal of droplets larger than 60 µm. Lower standards can be met by sophisticated, multi-chambered separators that incorporate coalescing plates and treat low flows (40 to 50 gpm) of a consistent influent concentration.
Hydrodynamic separators for municipal applications are not capable of removing stable emulsions or dissolved oil. OWSs are not usually applicable for general storm water runoff because by the time the oil reaches the device, it has emulsified or coated sediment in the runoff and is too difficult to separate.
Traditionally, 150-µm separation has been used, which typically results in an effluent oil and grease concentration of 50 to 60 mg/L. Flow-density-based separators are limited to removal of “medium”-sized droplets (100 to 140 µm) and have a low head requirement.
The most important characteristic affecting performance is oil droplet size, from which the critical rise rate can be determined. After determining the rise rate, design flow rate and effective horizontal separation area, the separator can be appropriately sized. The efficiency of separators is dependent on detention time in the chamber and on droplet size. When considering use of these systems for storm water treatment, landuse, site location and operation and maintenance should be taken into account.
A variety of hydrodynamic separators were investigated to provide performance details for use in selection criteria. Hydrodynamic separators have a smaller working area than conventional gravity separators and remove debris, sediment and surface grease and oil. For this project, specific details regarding 11 different proprietary systems were determined. This information will assist in the selection of a particular system depending upon the selection criteria. The data was obtained from the manufacturers’ specifications.
Additional information regarding hydrodynamic separators is provided in the full report. Manufacturers’ specifications for the systems reviewed are included as well as a U.S. EPA fact sheet for these systems.
The following criteria is recommended for UDOT purposes:
- Utilize the 60-µm-sized oil droplet for the basis of design for spill containment;
- Develop performance measures for oils and TSS; incorporate peak flows on small detention basins;
- Develop vendor selection criteria that includes TSS and oil droplet size;
- Include life-cycle maintenance costs;
- Evaluate a monitoring program and evaluation criteria;
- Incorporate TMDL-impaired waters in the selection process; and
- Incorporate West Nile virus concern.
Conclusions & recommendations
Current state of Utah and federal storm water discharge permits require the implementation of BMPs to reduce the discharge of pollutants to the maximum extent practicable (R317-8). Current state of Utah Division of Water Quality rules require storm-sewer discharges that discharge greater than 5 cfs into a receiving water to obtain a Stormwater Permit for Construction Activities and implement controls. These BMPs are considered for implementation to meet water-quality requiremewnts of the UDOT Phase 1 Municipal UPDES Permit conditions.
Storm water discharges from urban areas contain potential pollutants that can be characterized by land use activities and are largely dependent on climate patterns. Specific land-use analysis and evaluation of potential pollutants of concern is required prior to the design of treatment BMPs. Total suspended solids may be a target constituent of treatment BMPs, as the average concentration in local urban storm water is 116 mg/L, based on monitoring data by Salt Lake County, Salt Lake City and UDOT. The same data indicates that oil and grease concentrations in storm water flows to be typically in the range of 5 to 10 mg/L.
To remove sediment from storm water discharges, hydrodynamic separators may be considered to treat small storm events or the first flush produced during larger storm events. A design water-quality storm event of 1⁄3 of a two-year, 24-hour event (roughly 0.5 in.) has been identified as design criteria to size the water-quality control measure.
If hydrodynamic separators are chosen as the treatment control, additional design criteria include sediment storage capacity of the device, head loss, sediment particle size and overall efficiency of the treatment measure. Flows exceeding the design event will need to be routed around or bypass the treatment device. The hydrodynamic separator also provides some capacity to contain spills and litter (floatables) that occur within the basin or in hot-spot areas.
However, due to the high removal efficiency of solids, these treatment devices must be maintained on a regular basis. It is recommended that a strong maintenance program be implemented if these devices are utilized. It is also recommended that some monitoring be conducted on the units to document removal efficiencies and collection of floatables.