Saint Paul , Minnesota
St. Paul has been restoring wetlands since the 1980s to protect source water.
Background
The Saint Paul Regional Water Services (SPRWS) provides potable water to the cities of St. Paul, Arden Hills,
Falcon Heights, Lauderdale, Little Canada, Maplewood, Mendota, Mendota Heights, Roseville and West St. Paul.
The SPRWS treatment facility receives source water from a chain of reservoirs, which includes Pleasant Lake,
Sucker Lake and Vadnais Lake. These reservoirs are augmented with water from the Mississippi River and local
watersheds.

The SPRWS has been plagued with severe taste and odor problems in the water supplied to the public. Most of
these problems can be traced to the degraded water quality of the source water reservoirs brought on by excessive
algal growth.  In 1984 SPRWS  initiated an intensive diagnostic study of the source water reservoir system and
contributing watersheds. The results of the study indicated that nutrient enrichment, with phosphorus as a limiting
nutrient, had produced abundant algal populations, which in turn produced taste and odor compounds. Identified
sources of phosphorus loads into the reservoir system included the Mississippi River, internal loads from lake
sediments and local watersheds - primarily Lambert Creek.

Phosphorus loadings from the Mississippi River source and reservoir sediments were addressed by capital projects
funded by SPRWS under the state's Clean Water Partnership program. A Phase I  Clean Water Partnership (CWP)
diagnostic study of Lambert Creek was conducted under the sponsorship of the Vadnais Lake Area Water
Management Organization (VLAWMO) in 1991. This study revealed that wetlands along the course of Lambert
Creek had high sediment phosphorus concentrations and release rates.  The problem was further exacerbated by the
fact the wetlands had been drained when Lambert Creek became a county drainage ditch in the early 1900's and the
City of White Bear Lake had used the headwaters as a stabilization pond for the municipal wastewater effluents.
The subsequent wet/dry cycling of the wetland via climatological conditions provided for periods of decomposition
of detritis followed by flushing to Vadnais Lake. These findings were incorporated into the workplan, which called
for "reconstruction" of the wetlands via man made control structures on the outlets.

The diagnostic study and work plan were submitted to the Minnesota Pollution Control Agency (MPCA) for Phase
II CWP implementation funds.  The project was awarded Clean Water Partnership dollars to construct control
structures on the outlets of the three wetlands, Rice Lake, Grass Lake and Lambert Lake. With additional financial
assistance from the U.S. Fish and Wildlife Service and Ducks Unlimited Inc., construction was completed on the
                                                           structures at Rice Lake and Grass Lake in 1995.
   Lambert Creek Watershed
    SUCKER
   V&DNAIS
                v   o
        RICE L,
M3RASS L
                    LAMBERT L,
While it can be demonstrated that the Rice Lake
and Grass Lake implementations have been a
success, Lambert Creek continues to be the
dominant nutrient source in the eutrophication of
Vadnais Lake.  The long-term water quality goal
of 25 micrograms per liter P in Vadnais Lake
has not been met. During the open water
season, Lambert Creek can provide up to 55% of
the total P load to Vadnais Lake. Further
reductions in P loading are desirable for
improved water quality in Vadnais Lake.
Lambert Lake provides an opportunity, via
ponding, flow dispersion, and infiltration to
reduce the P loadings to Vadnais Lake.
Office of Water (4606M)
                                              816F100019
                                                                        January 2010

-------
Local Involvement
The VLAWMO and the SPRWS have an excellent track record in implementing sound water quality improvement
projects dating to the mid 1980's. All of the projects have been successful in phosphorus reduction and water
quality improvement, and each project has been completed on time and within the established budgets. The
Lambert Lake restoration project has a strong local support network with the VLAWMO, the SPRWS, and Ramsey
County as sponsors, and cooperation with the City of Vadnais Heights.  The project design represents a composite
of ideas, concerns, and suggestions from an ad hoc group of interested parties including the Vadnais Heights City
engineer, the Minnesota Department of Natural Resources (MN DNR) regional hydrologists, and the VLAWMO
and the SPRWS board members. The group has approved the plan and believes it will be both conducive to
implementation and successful in reducing phosphorus and improving the wetland hydrology.  The City of Vadnais
Heights is confident their concerns of property protection from flooding, improving the aesthetic value of the
wetland, and increasing the property value of the area residences will be addressed by the design.

Management Measures
SPRWS has completed phases I and II on Rice Lake and Grass Lake and has begun Phase III to address the
phosphorus contribution from Lambert Creek.  The primary focus of the work plan is to reduce the mass of
phosphorus runoff to Vadnais Lake. Long term statistical analysis has indicated that in order to lower taste and odor
episode frequencies to acceptable levels, total P values need to remain equal to or less than 25 micrograms per liter
in mixed layer samples from April through September.  This should reduce the frequency of chlorophyll A
concentrations exceeding 20 micrograms per liter, which is the nuisance level at which taste and odor episodes are
likely.  In order to achieve the 25 ppb management goal, Lambert Creek P loads need to be reduced by as much as
50%.

Lambert Creek's contribution to Vadnais Lake water quality degradation is significant.  As mentioned earlier,
Lambert Creek can contribute up to 55% of the April through September total P load while contributing only 10% of
the flow (1995 numbers).  Modeling efforts strongly indicate that significant water quality improvements to
Vadnais will not occur until there is a reduction in the P load from Lambert Creek. Lambert creek provides, due to
the room for improvement and relative small scale of the watershed, the best opportunity to improve water quality in
Vadnais Lake.
                                                                         It has been demonstrated that the
                                                                         previous work on Rice Lake and
                                                                         Grass Lake has reduced and
                                                                         stabilized phosphorus
                                                                         concentrations in the wetland
                                                                         outflows since 1995. Of
                                                                         particular interest is the cessation
                                                                         of phosphorus "spiking" in years
                                                                         following low precipitation years.
                                                                         The mean and standard error of
                                                                         the flow weighted mean
                                                                         concentrations of phosphorus
                                                                         dropped significantly after the
                                                                         wetland enhancements were
                                                                         completed. Critical homes were
                                                                         successfully protected in high
                                                                         flow conditions as a result of the
                                                                         berm installations as well.

                                                                         The SPRWS has permanent
automated monitoring stations at five locations within the Lambert Creek watershed, including Parshal flumes and
computerized equipment, and has collected continuous flow recordings for 17 years. The flow record and
subsequent analysis show that the average runoff coefficient for the Lambert Lake sub watershed is  extremely low at
0.04 indicating that substantial "catchment" and local "ponding" is available in this area.  Individual sub watershed
phosphorus discharge calculations have also shown Lambert Lake retains phosphorus in all but 2 years of the 17
year flow record; these were all in very high flow years 1993, and 1996.  In addition, the US Geological Survey and
Rice Lake Phosphc
Pre and Post Impoun
1200 -,^^^^^^^^^^— ^^^^^^^^^^—
1000
3"
Q.
Q. snn

E goo
O
.n
Q. •
to 400
o
.n
°" °00
ro
£ 0


I
1
1
1985 to 1994 1995 to 2000
-200 -
Period
Drus
dment
+ Mean + 1 SB Deviate
Mean '
Mean - 1 Stt Deviation


I
f

2000 to 2005

Office of Water (4606M)
                                               816F100019
January 2010

-------
the University of Minnesota sedimentigraphy investigations independently identified a layer of glacial sand that lies
immediately below the thin layer peat cover of the wetland (2 meters deep).  These sediments are remarkably
different from the upstream wetlands Grass Lake and Rice Lake that contain at least 30 feet of peat below the
surface. This glacial sand may play an important role in recharging the ground water aquifer.  The project design
takes advantage of these catchment and recharge properties by spreading the ditch flow through the wetland
promoting sheet flow and detention.

The goals of the plan include:
•   Purchasing easements
•   Creating an alternative ditch route
•   Abandoning the existing ditch by filling with excavation material
•   Construction of a pond and distribution weir
•   Construction of an emergency overflow weir
•   Distribute the ditch water to the wetland
•   Avoid any increase in flood levels

The project constraints are multiple due to the lack of stream slope, the proximity of homes, and the relatively flat
area surrounding the site. Additional constraints include the high cost of easements and construction. The following
plan was presented to the Vadnais Lake Water Management Organization's (VLAWMO) Board for review and later
approval. The plan incorporates ditch abandonment, excavation, berm construction, and weir construction, while
minimizing the purchase of easements and any increase in the current 100 year flood level. The plan also provides
greater flood protection for critical homes lacking the necessary freeboard elevation above the 100 year flood level.

The construction plan of Phase III consists of installing a 530 foot sheet pile weir with 45 adjustable weir openings.
A 1.7 acre pond was excavated to a depth of 6 feet to act as a stilling pool and the outflow will be dispersed into the
greater wetland.  The current ditch was filled in for a reach of 100 feet (County Ditch 14), and a 2226 foot
emergency overflow ditch was excavated to the east of the current ditch. A small berm will be installed along the
length of the filled portion of the ditch and the emergency over flow to prevent short circuiting of the overflow into
the filled ditch.  Access to the site was made available by the construction of a 2780 foot temporary access and a
permanent 600 foot class 5 utility road. The weir was designed for dispersion only with minimal impoundment and
head loss; about 0.1 feet of head. A pond outlet will be constructed at a proper elevation to direct flows greater than
100 cubic feet per second (cfs) to the emergency overflow route.  Flood protection will be provided to the critical
homes that lack proper freeboard by constructing a berm along the north portion of the excavated pond and the
construction of a graded swale to route the storm water away from adjacent homes to the pond/ditch system. The
construction was completed in February of 2005.
                           r»   v-
                           -   *»
                               so
                                *
Office of Water (4606M)
                                                816F100019
January 2010

-------
Contingency Planning
Analysis of the post construction monitoring data will reveal the level of success of the design.  In the event the
management goals are not sufficiently achieved, plans have been drawn to construct and operate an alum feed
system upstream of Lambert Lake at the Grass Lake weir. This system has been simulated in exhaustive jar testing
protocol and subsequent water quality modeling. The phosphorus removals are expected to reach above 90% using
8 ppm as aluminum as a constant dose rate. The proposed feed system would be very expensive to operate and
maintain, however.

Measuring Program Effectiveness
The St. Paul Regional Water Services will continue to maintain and operate the automated monitoring network. In
addition, water quality models have been in use relating watershed discharges to reservoir water quality. These
systems will be used to determine improvements in discharges from Lambert Creek and the resulting reservoir water
quality.

For further information contact:
Dave Schuler
Chief Engineer
St. Paul Regional Water Services
(651) 266-6286
dave. schuler@ci. stpaul. mn.us
Office of Water (4606M)                            816F100019                                    January 2010

-------