BMPs work! Stormwater runoff from the LID subdivision was much lower than from the traditional subdivision during construction. And, BMPs continued to work! After construction, stormwater runoff was unchanged from pre-development levels.

General Findings (Non-Technical)

Traditional Subdivision

What happened in the Traditional subdivision as a result of construction?

Before construction this area had a few small parking spaces, but it was mostly open with grass cover. As construction of the new subdivision began, the asphalt road, houses and driveways were installed. Due to these additions of impervious surfaces, the volume of water that ran off the site per unit area increased by 1550%! Imagine that the runoff before construction is the small cylinder below. As a result of construction, the stormwater volume running off the site increased to the size of the large cylinder!

Traditional subdivision construction runoff increase as a cylinder
Traditional runoff before construction vs. Traditional runoff due to construction: 1550% higher!

What happened in the Traditional subdivision after construction?

After all of the road surfaces, driveways and houses were installed, dramatic changes were found in stormwater runoff from the Traditional subdivision. The stormwater volume per unit area running off the site after construction increased by 1550%! Imagine that the runoff before construction is the small cylinder below (left). As a result of construction, the stormwater running off the site increased to the size of the large cylinder!

Traditional Subdivision post-construction runoff increase as a cylinder
Traditional runoff pre-construction vs. Traditional runoff post-construction: 1550% increase!


As a result of these large increases in stormwater running off the site, the amount of certain key pollutants that entered local waterways also increased. These pollutants, such as phosphorus, metals, and sediment, harm aquatic life and create nuisance algae blooms in local waterways.

LID Subdivision

What happened in the LID subdivision as a result of construction?

Before construction, this area was an old gravel pit, with a wetland in one section. As construction of the new subdivision began, practices were utilized to keep as much stormwater runoff on the site as possible. Due to these efforts, the stormwater runoff volume per unit area from the site during construction was 29% less than what ran off the undeveloped site! Imagine that the runoff before construction is the small cylinder below (left). As a result of LID construction practices, the stormwater running off the site decreased to the size of the smaller cylinder!

LID subdivision construction runoff decrease as a cylinder
LID runoff before construction vs. LID runoff due to construction: 29% decrease!

What happened in the LID subdivision after construction?

The stormwater runoff volume per unit area from the LID subdivision after construction was 42% less than what ran off the undeveloped site! Imagine that the runoff before construction is the small cylinder below (left). As a result of LID practices, the stormwater running off the site decreased to the size of the smaller cylinder!

LID subdivision post-construction runoff decrease as a cylinder
LID runoff pre-construction vs. LID runoff post-construction: 42% decrease!


Although there was a large decrease in stormwater runoff, the amount of a couple of pollutants leaving the site did increase. Sediment and phosphorus export increased, due to some fertilization by homeowners in the grassed swales and possible grass clippings. However, the increase was much less than the increase in the traditional subdivision.

Technical Results

Click on each box to learn more about the technical results of each area:

Water Quantity


Stormwater runoff volume increased (894%), as did runoff depth (1550%) from the traditional subdivision, based on the calibration period regressions. Peak discharge rate also increased by 400%. In contrast, as development occurred in the LID subdivision, adecrease in the volume of stormwater leaving the site was found (-97% for volume, -29% for depth). No change was found for peak discharge rate in the LID subdivision


The Traditional subdivision experienced increased volume, depth, and peak discharge of stormwater, while the LID subdivision experienced decreased volume, depth, and peak discharge.




Analysis of yearly survey results of all residents in the study watersheds indicated that, in general, residents of all three watersheds across several years of study showed no differences in behaviors such as lawn care (i.e., watering, fertilization), pet waste disposal and car washing.

Water Quality


Concentrations of most pollutants did not change in runoff from the traditional subdivision. This occurrence is likely due to the effective use of erosion and sediment controls. In contrast, concentrations of many pollutants in the LID subdivision increased, likely due to the instability of recently vegetated grass swales.


Concentrations of pollutants in the traditional subdivision either decreased or did not change after construction. However, in the LID subdivision, concentrations of some pollutants decreased, but others such as Total Suspended Solids (TSS), Nitrite + Nitrate Nitrogen (NO3-N), and Total Phosphorous (TP) increased.

TSS and TP mass export increased from the LID subdivision, but increases were much less than from the traditional subdivision.



Impervious Surface Change

No change in runoff was seen for the LID subdivision, but a logarithmic increase in stormflow was found from the Traditional subdivision, even though precipitation showed no trend.

Lag Time

Runoff lag times for the LID watershed were significantly greater than for the traditional watershed for small (< 25.4 mm) but not large (>4 mm) storms; short duration (< 4hr) but not long duration (> 4 hr) storms; and low antecedent moisture condition (AMC) (< 25.4mm) storms but not high AMC (> 25.4 mm) storms.


Three different types of driveway surfaces were investigated in this portion of the study: Ecostone® concrete block pavers, crushed stone and traditional asphalt. The crushed stone driveways had the least amount of runoff over the course of the 22 month study, with a mean of 0.04 mm. Runoff from the Ecostone® driveways was significantly higher at 0.5 mm, and runoff from the asphalt driveways was significantly higher than both of the other driveway types, at 1.8 mm. As would be expected, pollutant export was highest from the asphalt driveways





Cluster Design

The cluster design helped to reduce overall imperviousness, and increase open space. A cluster approach is recommended for future developments.

LID-based Regulations

Because many waivers were required for the construction of this project, an LID ordinance is recommended to facilitate adoption of this approach in other towns.

Disconnect Stormwater

The percentage of impervious coverage has been related to water quality and habitat impairments. As part of the low-impact development approach, disconnection of stormwater sources should be considered at the planning phase. Common disconnects include gutter design (to ensure that downspouts drain to pervious surfaces), road and driveway design. By emphasizing disconnected stormwater, the thresholds identified for water quality impacts from traditional developments may not apply.



The intensive one-on-one education methods that are traditionally thought to produce the most effective behavioral changes did not perform as intended. The involvement of a social scientist might be helpful to ascertain appropriate education methods.

Bioretention Maintenance

Maintenance of the cul-de-sac bioretention area resulted in excessive weed trimming with damage to shrubs. It is recommended that proper maintenance techniques be implemented to ensure proper function, aesthetics and plant health. Such techniques include maintenance of flow paths, surface water storage capacity and mulching to reduce weeds and the need for mowing.

Paver Maintenance

Due to the excessive loading of wind-blown fine particles loaded on to the pavers before construction was complete, the infiltrating ability of the roadway was reduced. Therefore, it is recommended that if a pervious surface has high loading of fine particles, the surface should be maintained by vacuum suction and replacement of the infill materials.

Turf Dam

The turf has begun to creep over the edge of the roadway in some places. Some consideration should be given to avoid this growth, as it can channel water down the roadway and prevent it from entering the swale.

Fire Hydrant

The fire hydrant on site was flushed at one point, and the infill material on the roadway was washed out in a large area. It is recommended that if this practice is necessary, care should be taken to direct the flush water away from the roadway and into an area that will not erode.

Seed Mix

A special turf seed mix was used at the site that had low fertilizer, water and maintenance requirements. However, several homeowners reseeded large areas with standard seed. Therefore, it is recommended that the use of the special seed mixture is specified in the homeowners’ association documents.

Further Study Recommendations

Groundwater Effects

Groundwater monitoring was not performed as part of this study. However, the addition of this type of monitoring would help to answer questions about the fate and transport of pollutants as a result of the LID infiltration practices used, versus traditional stormwater methods. 

Social Indicators of Behavior

The intensive education methods used in this study did not provide the expected results. Future watershed studies should include social scientists to better understand the role of humans in a watershed.

Testing of Soils

Some soil testing was performed as part of this study, however, a more intensive soil testing effort would provide valuable information on the fate and transport of pollutants. 

Economics of LID

Economists should be involved in LID watershed studies to appropriately assess LID costs and benefits compared to traditional development.



Soil compaction due to heavy equipment use of a temporary access road caused problems with infiltration in rain gardens and swales. Therefore, it is recommended that soil compaction be kept to a minimum, and that hard-surface roads are used for access during construction.

Undisturbed Soils

To help maintain the overall infiltration capacity of the soils on the site, it is recommended that as much of the site’s soils and vegetation as possible is left undisturbed.

On-site Supervision

Because LID practices are still fairly new and contractors are unfamiliar with their installation, it is recommended that a person versed in installation of LID techniques is on-site during construction. In addition, it is imperative that this person has the ability to make adjustments to the plan in the field, as necessary

Earthen Berm

The installation of the earthen berm to reduce export of sediment and runoff worked well, and is a recommended construction best management practice.

Grassed Swales

Due to the fact that grassed swales are vegetated systems and are vulnerable to erosion until stabilized, it is recommended that temporary erosion and sedimentation controls are implemented when swales are installed. See the Connecticut Guidelines for Soil Erosion and Sediment Control for specific recommendations. 

Soil Testing

Contractors are accustomed to applying a starter fertilizer to loam when seeding a lawn. This practice occurred at the site, despite the fact that soil tests showed that no fertilizer was needed. Therefore, it is recommended that a soil test is performed when loam is applied to swales or lawns, and that the contractor follows the recommendations. More broadly, soil tests do not include a test for nitrogen, so even if homeowners or contractors have a soil test done, recommendations for nitrogen application are not provided. It is recommended that a nitrogen test be a component of standard soil testing.



The use of the control watershed was critical in this study to attribute the changes observed to the construction and management practices used, without bias from climatic variations. Therefore, the use of a similar control is advised in future monitoring efforts.

Forested Control

The control used in this project was a previously developed residential area. To make direct comparisons with the pollutant export from undeveloped areas, it would be helpful in future studies to include an undeveloped control with strong protection measures to ensure undeveloped status during the duration of future study. 

Sampling Methods

Although the continuous automated sampling used in the project was highly successful, relatively few event-based grab samples were taken. Grab sampling is weather dependent, and is difficult to perform at a remote site. Future projects should consider the proximity of the site if grab sampling is desired. In addition, a local volunteer could be enlisted to aid in collecting samples on evenings or weekends.

Electric Power

Solar panels were used to supply power at various times through the project. However, the ability to connect to the power grid provided more stability and gave more flexibility for sample preservation (i.e., the use of small refrigerators). Future projects should consider connections to the grid when possible.

Myths Refuted

Winter Infiltration

Although no formal seasonal analysis was performed to investigate BMP performance during winter months, paired watershed analysis of the LID subdivision indicated decreases in the runoff volume and rate as compared to pre-development. In addition, field observations indicated that the rain gardens, the grassed swales and the paver road functioned as designed through the winter months. 

Frost Heaving of Pavers

Frequent concerns are voiced about paver blocks heaving in the winter due to frost penetration. The EcoStone® paver roadway has remained in excellent condition, with no frost heaving noted. The installation of a proper bedding material and storage layer to encourage rapid infiltration into the subgrade is critical to this function.


Typically, intensive, one-on-one education methods are assumed to be the most effective means to bring about changes in attitude and behavior. However, on this project, this method did not bring about the desired behavior changes.