It's raining, but most people aren't giving the weather a second thought-they have already cancelled outdoor plans and made sure their windows are closed.  It's quite a different story for PB engineers in Baltimore, who are preparing to collect storm water runoff from construction sites and determine its sediment concentrations. 

We have two sediment load monitoring projects underway to determine how storm water from construction sites affects receiving streams.  The first began in February 2005 and focuses on two of three streams crossed by the MD 43 Extension project.  The second, which got underway in the first quarter of 2006, involves monitoring several streams that will be impacted by the construction of another roadway, the Hampstead Bypass.  Both studies were initiated by Maryland's State Highway Administration.

Maryland Route 43 Extension Study

The MD 43 Extension project will extend Whitemarsh Boulevard (MD 43) in Baltimore County.  The alignment crosses several streams and wetlands including Whitemarsh Run, Windlass Run, and a tributary to Windlass Run.  Bridges over Whitemarsh Run and Windlass Run have been constructed, as has a triple-cell culvert for the tributary.  Since the summer of 2003 we have been studying the long-term impacts that these crossings and development of the adjacent wetlands are having on the stability of these streams.

In late 2004, when construction was underway, we were contracted by the State Highways Administration's Highway Hydraulics Division to develop and implement the study that is the subject of this article.  The goal was to determine if the project was in compliance with National Pollutant Discharge Elimination System (NPDES) Storm Water Permits to control construction site runoff and post construction storm water management. 

NPDES regulations require carefully installed and maintained erosion and sediment control measures be implemented for all projects that disturb more than 465 m2 (5,000 square feet).  These measures, which include sediment basins, silt fences and diversion ditches and berms, must be installed at the initiation of construction and maintained until permanent soil stabilization has been completed. 

Our goals were to monitor the efficiency of erosion and sediment control methods in removing suspended sediment and to quantify the resulting sediment loading on receiving streams from construction grading during storm events.  More specifically, our tasks were to:

• Observe various phases of construction activities and document the effectiveness of commonly used erosion and sediment control practices at removing sediment from construction runoff
• Measure the effectiveness of sediment basins in collecting suspended sediment before discharging to receiving waters after storm events
• Determine the impact that sediment leaving construction sites had on the total suspended sediment (TSS) concentration and turbidity level of receiving streams during and after storms
• Help the Highway Hydraulics Division to develop a potential incentive/disincentive program.

Study Setup and Challenges

Four sediment basins were chosen to monitor the MD 43 construction project (Figure 1 on the following page).  Two discharge to Windlass Run, and two discharge to the tributary to Windlass Run. 

As with any environmental study, data collection has to be done in a way that isolates outside factors that could influence the results.  In developing this study, we were careful to isolate the sediment basins impact on the receiving streams by strategically locating sampling locations upstream, in between, and downstream of where the sediment basins' discharge reaches the streams.  Each basin had grab samples taken from the middle of the pond and at the outfall of the pond.

Technology Used

Initial setup was geared towards post-storm collections because equipment to sample during the storms was not available at the time.  After a significant rain event (nearly 1 inch), engineers visited the site and collected grab samples (Figure 2).

As the study progressed, portable automated samplers were installed that enable us to collect samples at the basin outfalls during a storm event.  These samplers monitor level and flow from the basins and collect runoff at pre-programmed intervals when flow depth levels reach a predetermined enable height that indicates a rain event of desired magnitude is occurring.  Each sampler collects 24 samples in 500 mL bottles.

A rain gage was installed on site and connected to one sampler along with a cellular modem (Figure 3).  This equipment enables us to download the information from that sampler remotely so we can determine the amount of rainfall and whether the discharge levels from the basin were high enough to reach the enable height for that sampler.  This information helps us to determine if a site visit is needed, thereby increasing the efficiency of the sampling team and minimizing "false alarms." 

Current Sampling Procedures

We monitor weather on a daily basis using several Web sites to determine when the next significant storm will hit and monitoring activities should commence.  Maintenance is performed on a bi-weekly basis to ensure the samplers are functioning correctly and that they are set to collect samples when the next storm occurs. 

The day after a storm, we use the remotely gathered data to see the quantity of precipitation and status of the sampler.  If the bottles were filled with sediment-laden water, then the sampling team visits the site to collect them from the four samplers.  While on site, we also collect bulk samples at each basin and in the streams.  All samples are taken to the laboratory and analyzed for TSS concentration and turbidity levels.

Findings So Far

During the construction phases, sediment levels were generally higher in the basins than in the discharge.  Streams showed elevated levels of sediment as they moved downstream of the project, but levels remained low as the drainage area is mostly undisturbed.  In general, sediment is contained in the ponds, and higher sediment levels can be contributed to large unstabilized portions of the drainage area. 

As the study has progressed, construction activities have mostly ceased around Windlass Run and sediment loadings in the stream and surrounding sediment basins have decreased accordingly.  It is expected that sediment loads will continue to decrease as construction nears completion.  The future of this project may hold additional sampling for other pollutants washing off the newly opened highway.

Throughout sediment sampling and monitoring activities, the riser and weir outfall structures were often not efficient at reducing TSS and turbidity.  Performance of both outfall structure types was inconsistent and occasionally the structures were the source of additional sediment loading.

The streams are influenced by the discharge from the ponds but are also adversely affected by disturbed areas not within the drainage areas of the four ponds.  These areas are intended to be intercepted by super silt fencing prior to flow into the streams. 

Lessons Learned

It is difficult to sample storm water consistently.  Weather predictions vary greatly and often it is not possible to know if a rain event was sufficient to merit sampling until after the storm has finished.  Automatic samplers have greatly increased the chances of collecting sediment laden water during a storm event but they are not failsafe.  Samples have been collected from several storms using the automatic samplers, but rarely do all basins discharge enough runoff to trigger all four samplers correctly for the same storm event.  Construction activities, drawdown times, and cold weather contribute to falsely triggering a sampler.  The correct trigger level that should be set for each sampler is always changing.

Even though the false triggers make the sampling schedule difficult to predict, we collected and analyzed significant data that made it evident that sediment control devices that appear to be constructed properly cannot be expected to work efficiently throughout the project duration without proper maintenance.  If maintenance is not performed prior to storms events, then the release of sediment to receiving waters is likely.

Specifically, the sediment basins often appear to be in good condition but clogged filter material and lack of dredging will drastically decrease their efficiency.  The result is that sediment laden runoff produced in small storm events, which should remain in the basin for a minimum of two days, frequently overtops the outfall structure.  If the water does not stay in the pond long enough, then the sediment will be released into the waterways.

Hampton Bypass Test is Underway

PB was asked recently by the State Highway Administration to conduct another water quality monitoring task on a project for which construction had not yet started-the Hampstead Bypass in Carroll County, Maryland.  The goal is to monitor storm water runoff quality before, during, and after highway construction.  At the time of writing, eight automated samplers have been installed at strategic locations to collect pre-construction runoff data.  It will be interesting to watch how sediment levels change during construction on the Hampstead Bypass.  Starting the task pre-construction will:

• Offer much more baseline data about the current conditions of streams in the area
• Show the total sediment loading for the life of this project. 

The data may also allow us to compare pre- and post-sediment levels to determine if sediment loading occurs only during the construction phases or if the constructed area has a longer lasting detrimental affect on the receiving waters.