Planting crew in cell 1, planting cattail in mid-June.

Introduction
Talkeetna, Alaska is a small unincorporated community of approximately 350 year-around residents and is part of the Matanuska-Susitna Borough located in South Central Alaska. It is about 115 road miles north of Anchorage and about 14 miles off the Parks Highway at the end of the Talkeetna Spur Road. In 1989, a public sewer system and two-stage lagoon facility were constructed to serve the core area of the community. At the time of the proposed improvements described in this report, the system consisted of a force main sewage collection system which delivered the collected effluent to a lagoon system for treatment and final disposal. The treatment facility consisted of two holding cells and a percolation cell. Approximately 40,000 gallons per day of sewage was pumped to the lagoon facility located adjacent to a slough on the Talkeetna River and disposed of through the percolation pit. The stored effluent in the holding cells was typically transferred to the percolation pit twice a year - just after spring breakup and before the onset of winter. Nearly 9 million gallons of effluent percolated as a slug flow through a thin layer of soil into the shallow ground water table.

Looking downstream from cell 1, 30 days after planting.

In October 2001 the Borough was given a Notice of Violation by the Alaska Department of Environmental Conservation (ADEC) for surpassing water quality discharge standards. Subsequently the Matanuska-Susitana Borough let an RFP for an upgrade to the system. The upgrade to the town’s water and sewer system was funded by the USDA and administered by the Matanuska-Susitna Borough.

The entire project included improving and modifying the existing lagoon flow control system; converting the percolation pit into a third lagoon; installing a new lift station and modifying two existing lift stations; building a constructed wetland treatment system; and camera inspection and cleaning of the existing force main. In addition, a previously non-existant SCADA monitoring system was installed for remote monitoring and control of all pumps and controls in the lift stations. The prime contractor was Construction Unlimited and the landscaping contractor was Evergreen Landscaping, both based in Anchorage, Alaska.

After a competitive bid process, a group consisting of Bob Gilfilian, P.E., of Gilfilian Engineering, Dr. Dave Maddux of Applied Wetlands Technology and Mark Sherman, P.E., of ASCG, Inc. were awarded a contract to design and oversee construction of a surface-flow constructed wetland. This system was a replacement for the percolation pit in the final treatment of effluent before its discharge to the receiving waters of the Talkeetna Slough. The following discussion deals only with the constructed wetland portion of the project.

Benefits Of Using A Constructed Wetland
Once the DEC issued a Notice of Violation, it was critical that the problems be remedied as soon as possible. Since Talkeetna is situated in a pristine environment and is considered by many to be the “Gateway to Denali”, the town wanted a system that mirrored the natural setting found around the community at the foot of Mt. Denali. Another consideration was the cost of operating and maintaining improvements to the current system.

Selection of a constructed wetland treatment system provided several benefits. When construction was completed, operation and maintenance costs were estimated at $2,000 per year; effluent discharge would meet DEC standards; and the treatment system created an attractive green space that complimented its natural environment.

The construction cost of the constructed wetland portion was approximately $440,000 which equates to $13 per square foot of constructed wetland surface area. This facility is capable of treating 14.6 million gallons of effluent per 145-day treatment season which easily meets the current needs of the town and allows for future expansion.

Looking from cell 1 at cattail after two months of effluent flow.

Discharge Water Quality Requirements
Prior to construction of the new lagoon and constructed wetland, this facility discharged effluent to a groundwater source and not a surface-water source. As such, water quality criteria was different for percolation pit disposal methods. The Notice of Violation given to the Borough prior to new construction related to sludge buildup in the lagoons, percolation cell performance and groundwater quality issues. Switching to a surface-water discharge was intended to alleviate those issues.

The ADEC sets the discharge permits for treated sewage wastewater discharge to receiving waters in Alaska. There are three main requirements; 1) reduction of fecal coliforms to a 30-day average of 20 cfu/100 ml, 2) 5-day BOD concentration maximum of 65 mg/L and 3) a total suspended solids maximum of 70 mg/L. Dissolved oxygen has a minimum of 7 mg/L and a maximum of 17 mg/L.

Cells 1 and 2 after heavy rain mid-June.

Climatic Conditions Are A Challenge In Alaska
A problem that Alaska faces in using constructed wetlands for wastewater treatment is the shortened treatment season caused by long, cold winters with deep frost penetration. The average winter temperature in Talkeetna is +10.4 degrees F° (-12 degrees C°) and winter darkness lasts for approximately seven months. This results in a maximum treatment season of 145 days. Ideally, a seasonal storage capacity large enough to hold an entire years worth of effluent will provide a safety margin to allow this type of system to work in an arctic environment.

Constructed Wetland Design Overview
The existing side-by-side lagoons had surface areas of 2.2 acres each and a working volume of 3.725 million gallons each. The conversion of the percolation pit to a holding lagoon increased the surface area by 1.1 acres and the holding volume by 1.935 million gallons for a total holding volume of 9.385 million gallons.

Once the effluent is delivered to the treatment facility, the flow process is a simple gravity-fed discharge from one lagoon to the next. The pumped effluent flows into lagoon #1 and travels the length of the lagoon before being discharged through gravity flow to lagoon #2. The effluent then travels the entire length of lagoon #2 before being discharged through gravity flow into lagoon #3. Finally, the effluent is discharged seasonally to the constructed wetland through a buried, four-inch HDPE line by opening mechanical distribution header valves.

The constructed wetland is a continuous system comprised of six cells with an operating depth of 12 inches. Each cell is separated from the following cell by a deep water trench that is 4 feet deep and acts as a flow redistribution zone. This zone allows for remixing of the effluent, maximizing an even flow across the entire width of the cell and minimizing the potential for channels (short circuiting) developing through the wetland.

This free-water surface flow system has a surface area of 35,000 square feet and a volume of 2,618,000 gallons. Current flow-through is 395 m3/day (105,000 gpd) with a theoretical hydraulic detention time of 1.86 days. Discharged effluent flows through a V-notched weir to an 8” HDPE buried line that discharges to the Talkeetna River slough, an anadromous fish stream.

Plant Selection
Macrophytes chosen for planting in the constructed wetland were based on four criteria. They had to be indigenous to the area, have an ability to colonize rapidly, be able to withstand high pollutant loads of ammonia and provide a large surface area for periphyton attachment. Based on these criteria, five species were selected for the project. Cell 1 was planted with Typha latifolia (broad-leafed cattail), cell 2 with Scirpus validus (soft-stemmed bulrush), cell 3 with Carex utricularia (common sedge), cell 4 with Calla palustris (calla lily), cell 5 with Carex aqautilis (blue-green sedge), and cell 6 with Carex utricularia (common sedge).

It was expected that cells 1 and 2 would be subjected to the highest loads of ammonia. Cattails and bulrush were placed in these cells because both species have shown a remarkable ability to thrive while withstanding high pollutant loads that include ammonia. These species also produce a significant volume of biomass which provides an available carbon source to microbes in the substrate.

Placement of the common sedge in cell 3 was based on the known ability of the sedge to rapidly colonize the open areas and provide a large surface area for periphyton attachment. Both species of sedge used in this constructed wetland do well in polluted water of medium to low concentrations. Cell 3 is far enough along the treatment train that any high concentrations of ammonia should be reduced to levels that will not adversely affect the plants.

Cell 4 was planted with calla lily, which is primarily a floating plant. Calla is slightly rooted at its base in the substrate and sends out thick, floating stems sometimes ten feet in length, with large leaves. The remarkable attribute of this plant is the huge amount of subsurface aerial root production along each stem. This species has more biomass in the roots floating in the water column than in the leaves and stems above the water line. Not only does the plant supply a thick mat of vegetation below the water surface to filter out suspended solids, it also produces a large volume of leaves that decompose rapidly upon senescence and add to the detrital, vegetative mat. Furthermore, since it spreads out and covers the entire water surface, it minimizes the oxygen transfer between the atmosphere and the water surface, providing a near perfect environment for denitrification to occur.

Cells 5 and 6 were planted with blue-green sedge and common sedge respectively, both of which can be found growing along the banks of the receiving water. These two species will grow thick, luxuriant stands of plants, which will effectively filter out seeds and vegetative propagules from the cattail, bulrush and calla lily. Although these three species are indigenous to the state, they are not indigenous to the local area and spreading of the species by accident is not desired. These two species of sedge provide final polishing of the effluent before its discharge to the receiving waters.

Site Preparation For The Constructed Wetland
Before construction, the site immediately surrounding the lagoons was a flood plain forest consisting of birch, alder and spruce trees with a thick tangle of sword ferns, devils club and a variety of grasses. All vegetation was stripped adjacent to the existing lagoon structures and a trench was excavated with a bottom width of 32 feet, a length of 1,000 feet and side slopes of 3:1. After the grade was established and compaction achieved by roller to 95%, a non-woven, needle punched polypropylene fabric was placed on top of the compacted soil. A 30-mil polypropylene liner was placed on top of that, then a final layer of geotextile was laid down.

A layer of top soil 12-inches thick was placed on the liners on the bottom and side slopes of the cells. The soil was manufactured to a minimum specification organic content of 10% and a maximum of 20% by dry weight of finely chopped, well mixed organic materials. To keep clumps and rocks out of the mix the non-organic portion of the planting soil was required to be less than 2 mm in size.

Once topsoil was in place the planting crew began laying out the various planting grids and then using hand-held pluggers to remove soil plugs for planting purposes. Spacing for the various plants was either 12-inch, 18-inch or 24-inch centers depending on the species being used.

Plants And Planting Process
All the plants except calla lily were purchased from a wetland nursery in Montana. The calla was harvested from a lake in Fairbanks and transported by truck to the site the same day they were harvested. The other plants arrived by air and were held until they were needed in a pool designed specifically for wetland plants. Both the cattail and bulrush were planted on 18-inch centers requiring approximately 1,760 individual plants of each species. Both species of sedge were planted on 12-inch centers, requiring approximately 11,800 plants and the calla lily was planted on 24-inch centers requiring approximately 984 plants.

The planting process was the same for all plants except the calla lily. For the cattail, bulrush and sedge, a hole was punched in the soil, the soil plug was removed, then approximately 2 cups of diluted Humi-Zyme®, a liquid fertilizer, was poured into the hole. The plant plug was inserted into the hole, tamped down with some soil and a final watering of diluted liquid fertilizer was applied. Since the calla lily was harvested by hand from a wild stock, it was not rooted in a tube or pot. Each plant consisted of rhizome-like stalks approximately 18 inches to 24 inches long with substantial aerial roots along each rhizome. These plants were simply placed on the substrate with a shovel-load of soil placed on the end of the rhizome structure and then tamped down by foot.

After each cell was planted it was thoroughly watered but not flooded. Once all the plants were in, the cells were flooded to a depth of 1 inch for 10 days, then the flooding depth was raised to 6 inches for 20 days. The depth was then raised to the operating depth of 12 inches for the following 30 days. No effluent flowed into the constructed wetland during this 60-day acclimation period.

Project Timeline
Construction began on the project in the spring of 2003 with the macrophyte planting taking place in mid-June 2003. Following a 60-day transplant acclimation period, effluent flow through the constructed wetland began in mid-August 2003. Initial reduction of pollutants was exceptional. Following a six-week discharge period the flow was shut down for the winter. The following year was the first full treatment season for the constructed wetland and, as can be expected for initial operating seasons, water quality results were mixed.

Discharge structure from lagoon to head of cw.

Water Quality Results For The First Operating Season
The three main discharge parameters the system is required to meet are total suspended solids (TSS) of 70 mg/L, biological oxygen demand (BOD) of 65 mg/L and fecal coliforms (FC) of 20 cfu/100 ml. The first two parameters were well within the required reduction rates for all sampling events during 2004, however the FC met discharge standards only once out of four sampling events during 2004.

The effluent exiting the lagoons to the constructed wetland inlet was low in TSS and BOD and the discharge concentrations to the receiving waters averaged well below the ADEC discharge standards. Fecal coliform concentrations exiting the lagoon to the constructed wetland inlet were not as high as was expected except for two sampling events in the fall when concentrations were over 28,000 cfu/100 ml. The FC discharge concentrations to the receiving waters for three sampling events in 2004 were well above the desired ADEC water quality standards.

Outlet of cw to receiving water at full flow.

Problem Areas
Constructed wetland systems usually require a large volume of plants for initial planting purposes. However, this system is small compared to some systems in the lower-48 and only 16,300 plants were required. This still presented a problem because Alaska has no commercial wetland nurseries and although commercial growers in the state were contacted two years prior to the project, none expressed an interest in the wetland plant nursery business. Therefore, most of the plants were purchased from a wetland plant nursery in Montana and air-freighted to Alaska which added considerable cost to the planting of the constructed wetland.

Another problem which was not anticipated was migration of one of the species due to wind. The calla lily has large leaves which protrude like a flag on the surface of the water. Three days after flooding the cells to 12 inches, a stiff wind caught the large leaves like sails and pulled the plants from their mooring, blowing most of them into the upwind end of the cell. This left a large amount of open water available to waterfowl, which can have the effect of increasing the fecal coliform and BOD concentrations in the water.

A drawback from the client’s point of view is that constructed wetlands do not show their full potential for pollutant removal until about three years of growth have passed. Unlike engineered systems which are expected to work as soon as they are turned on, biological systems take time to mature. Full plant colonization that allows the various plant and microbe communities to establish themselves and fill all the required niches for efficient pollutant removal takes several seasons.

Lessons Learned
One of the problems this project faced was the very short timeline for construction and implementation of the constructed wetland system. Even a delay of two weeks can spell disaster for a project here in Alaska because the construction season is so short. Allowing a newly established biological system a season of growth before it receives effluent is ideal. However, this project was under pressure to complete planting the constructed wetland in time for it to be functional and ready to receive effluent the same construction season. Therefore, the minimum time allowable for establishment of the plants prior to effluent flow was determined to be 60 days. By mid-August that establishment period had been reached and effluent flow began.

Part of the performance of the landscaping effort was a guarantee that 90% of each species would survive transplant for a period of 60 days. For the most part this occurred, but two areas of plantings for one species in particular was adversely affected and had to be replanted by the contractor. As of the termination of the 2004 season those replanted areas still had not reestablished themselves. Either that particular species had a problem with the seed stock used by the greenhouse or those areas where the planting did not survive had soil that was toxic in some way. Planting procedures followed by the contractor were identical to the successful plantings, so no fault could be found with the contractor.

Acquiring all the necessary permits was somewhat complex because the Talkeetna River Slough is designated an anadromous fish stream. As such, the Alaska Department of Fish and Game became involved in the permitting process, along with the ADEC and the Army Corps of Engineers. Coordination and negotiation with all the agencies involved caused some problems which may have been avoided with more comprehensive permit planning.

Under a competitive bid process we were unable to select the contractor of our choice and there was some concern regarding unknown performance. Additionally, this was the first commercial constructed wetland built in Alaska and therefore no Alaskan contractor had specific experience with these systems. However, we worked with an excellent general contractor, construction crew and landscape contractor, all who made a potentially difficult project into a straightforward one. The construction schedule for this project worked well and there were very few changes I would make for a similar project.

For more information contact Dave Maddux, Ph.D., Wetland Ecologist, Applied Wetlands Technology, P.O. Box 81091, Fairbanks, Alaska 99709, e-mail: davemaddux@wetlandsoptions.com.