HomeMy WebLinkAboutLake-Cornelia_Edina-UAA-Alum-Feasibility-Study_2019
4300 MarketPointe Drive, Suite 200
Minneapolis, MN 55435
952.832.2600
www.barr.com
Alum Treatment of Lake Cornelia
Feasibility Analysis
Prepared for
Nine Mile Creek Watershed District
July 2019
P:\Mpls\23 MN\27\2327634\WorkFiles\Lake Cornelia\Alum Treatment and Lake Sediment\Feasbility Study\AlumTreatFeasStudy_July
2019_071519_changestracked.docx
i
Alum Treatment of Lake Cornelia
Feasibility Analysis
July 2019
Contents
1 Introduction ................................................................................................................................................................................. 1
2 Internal Loading Background ................................................................................................................................................ 3
2.1.1 Internal Loading in Lakes ......................................................................................................................... 3
2.1.2 Mixing Effects of Benthivorous (Bottom-feeding) Fish ................................................................ 3
3 Sediment Coring ........................................................................................................................................................................ 4
4 Alum Application and Dosing Plan ..................................................................................................................................... 6
5 Cost Estimate ............................................................................................................................................................................... 8
5.1 Opinion of Probable Cost ........................................................................................................................................ 8
5.2 Cost-Benefit Analysis ................................................................................................................................................. 8
6 Site Access ..................................................................................................................................................................................10
7 Permitting ...................................................................................................................................................................................12
8 Timing ...........................................................................................................................................................................................13
References .............................................................................................................................................................................................14
ii
List of Tables
Table 1 Average phosphorus concentrations in sediment cores collected in Lake Cornelia in 2008
and in 2019 ............................................................................................................................................................ 4
Table 2 Recommended alum and sodium aluminate application dose for 2019 and the total
application dose after a second application ............................................................................................ 7
Table 3 Engineer’s opinion of probable cost ........................................................................................................... 8
Table 4 Summary of annualized cost per pound of total phosphorus removed from conducting an
alum treatment in Lake Cornelia ................................................................................................................... 9
Table 5 Summary of annualized cost per unit reduction (µg/L) in summer-average total
phosphorus concentration………………………………………………………………………………………………. 10
List of Figures
Figure 1 Sediment Coring Locations ............................................................................................................................. 5
Figure 2 Potential Staging Locations ......................................................................................................................... 11
1
1 Introduction
The Nine Mile Creek Watershed District (NMCWD) recently completed a Use Attainability Analysis (UAA)
for Lake Cornelia (updated from 2010) and Lake Edina (first version) to assess the water quality and
prescribe management activities to improve lake water quality (see Barr Engineering Co., 2019). One of
the recommendations of the study was to conduct an alum treatment of Lake Cornelia (North and South
Basins) to bind (or immobilize) the phosphorus in lake bottom sediments and prevent release of the
phosphorus into the water column during summer months. All lakes accumulate phosphorus (and other
nutrients) in the sediments from the settling of particles and dead algae. This reservoir of phosphorus can
be reintroduced or recycled to the lake water and be used by plants and algae for growth. The recycling
of nutrients from the sediments to the lake water is known as “internal loading”. The process of internal
loading is complex and can be affected by temperature, oxygen, pH, wind mixing, and bioturbation by
organisms such as bottom-feeding fish species.
Significant reductions in internal phosphorus loading in Lake Cornelia are an integral part of the overall
strategy to reduce phosphorus concentrations as well as the magnitude and frequency of algal blooms in
Lake Cornelia. The lake water quality modeling conducted as part of the UAA demonstrated that internal
release of phosphorus from lake bottom sediments is a significant source of phosphorus to Lake Cornelia,
contributing an estimated 14%-40% of the annual phosphorus loading to North Cornelia during modeled
years, and an estimated 14%-19% of the annual phosphorus loading to South Cornelia. The lake water
quality modeling also demonstrated that an alum treatment will be highly effective in reducing the
summer-average phosphorus concentration in Lake Cornelia. Based on these findings, we recommend
conducting an alum treatment of Lake Cornelia (north and south basin).
Internal loading often occurs during time periods when flows into and out of Lake Cornelia are lower, and
the phosphorus tends to remain in the lake water column (i.e., is not flushed out). Since Lake Cornelia is
quite shallow, the effects of these internal loads are enhanced compared to deeper lakes. One of the
reasons to conduct an alum treatment now rather than after the other management activities have been
implemented is to improve summertime clarity and promote native aquatic plant growth. The City of
Edina is currently engaged in an effort to control invasive curly-leaf pondweed by treating with endothall
in the spring. Alum treatment is expected to promote aquatic plant community growth and subsequently
promote competition with curly-leaf pondweed. We expect that the alum treatment will enhance the
overall curly-leaf pondweed control efforts.
This study provides the results of sediment coring and phosphorus fraction analysis, alum (aluminum)
dosing, cost-benefit analysis of the treatment, site access, permitting requirements, and timing/schedule
of the application for Lake Cornelia. As part of this effort, the potential benefit of treating Swimming Pool
Pond with alum was evaluated using findings of a study conducted by University of Minnesota Saint
Anthony Falls Laboratory (Natarajan and Gulliver, 2019). Using sediment collected from the bottom of
Swimming Pool Pond, the study showed a low internal loading rate. This was confirmed with phosphorus
measurements within the pond. While there is some internal loading in Swimming Pool Pond, the benefit
of treating the pond with alum is expected to be minimal.
2
The sediment core analysis and lake water quality modeling conducted as part of the UAA demonstrated
that internal loading is not a major source of phosphorus to the Lake Edina. Therefore, we do not
recommend an alum treatment on Lake Edina at this time.
3
2 Internal Loading Background
2.1.1 Internal Loading in Lakes
Internal loading is a natural process in lakes and is mainly driven by temperature and oxygen availability.
Traditionally, internal loading has been attributed to deep lakes with warm surface waters and cooler deep
waters that mix infrequently and develop low oxygen conditions in bottom waters. In shallow lakes like
Lake Cornelia, stable, summer-long stratification due to temperature differences does not usually occur.
Instead, weak temperature gradients are formed during daytime hours and calm periods. This weak
stratification can cause small microzones or pockets of low oxygen just above the sediment surface. These
microzones are difficult to detect during periodic sampling because they are usually limited to a small
water layer above the sediment and exist for short periods. Oxygen depletion over-night can be
significant as algae respire (use oxygen) during this period instead of producing oxygen as they do during
daylight hours. Phosphorus is released from the sediment under these low oxygen conditions and since
the lake is shallow this phosphorus is immediately available for uptake by algae and plants the following
day. In shallower areas, algae will actually grow on the sediment surface and directly uptake nutrients
from the sediment. In some cases, pH can increase due to algae and macrophyte growth and phosphorus
can be released from the sediment if pH is high enough (greater than 9.5) for an extended period.
Phosphorus released from the sediment through internal loading processes is considered immediately
available because it is in a dissolved form that algae and plants can use directly. Watershed phosphorus
loading is generally 35-45% dissolved (on average in Minnesota) while the remaining portion is in the
form of particles (either soil or organic matter) that become part of lake sediment. The particulate form of
phosphorus cannot be directly used by algae or plants until it is released from the particles or organic
matter.
Phosphorus taken up by organisms in lakes (including algae and plants) is returned to the sediment when
the organisms die. Once in the sediment, much of this phosphorus can then be released again after the
organic matter breaks down, continuing the internal loading process. While this is a natural process in all
lakes, additional inputs of phosphorus due to human activity have caused increases in both the total
amount and the rate of internal phosphorus loading in lakes.
2.1.2 Mixing Effects of Benthivorous (Bottom-feeding) Fish
A 2018 fishery survey of Lake Cornelia identified a large population of benthivorous (bottom-feeding) fish,
specifically black bullhead and goldfish. Benthivorous fish species increase mixing of lake sediments,
which in turn can increase the rate (or speed) of internal loading in lakes. In addition, species like black
bullhead and goldfish can actually increase the depth of sediment mixing in lakes. This means that a
greater amount of phosphorus can be transported from the sediment to the water. Because there is little
to no oxygen in lake sediment just below the sediment water interface, the pool of phosphorus that might
not be available under ‘normal’ conditions without bottom-feeding species is physically pushed out of the
sediment due to mixing. It should be noted that benthivorous fish do not cause internal loading but they
may increase that rate of internal loading and they may increase transport of phosphorus from bottom to
surface waters.
4
3 Sediment Coring
Sediment cores were collected to determine the different fractions of phosphorus in the sediment (i.e.
phosphorus fractionation). Some forms of sediment phosphorus contribute to internal loading while
others remain in the sediment. Conducting phosphorus fractionation of the lake-specific sediments helps
maximize effectiveness and longevity of an alum treatment through development of a targeted
application and dosing plan. The sediment cores were analyzed for mobile phosphorus (mobile-P) (iron-
bound and loosely-sorbed phosphorus), aluminum-bound phosphorus, organically-bound phosphorus,
and calcium-bound phosphorus (calcium-P). The sediment was also analyzed for percent water and
percent organic carbon to calculate the density of the sediment. Mobile phosphorus is the phosphorus
fraction that that causes internal loading during the summer months when oxygen is low. This fraction is
also used to calculate alum (aluminum) doses necessary to prevent internal loading. Organically-bound
phosphorus slowly decays over time and this fraction is also important as it will eventually become mobile
phosphorus and lead to future internal loading. Aluminum-bound and calcium-bound phosphorus also
provide information about sediment mixing, the layer of sediment actively releasing phosphorus, and
historical phosphorus inputs.
A total of five sediment cores were collected from the lake bottom in 2008 and five cores were collected
in June 2019 (see Figure 1). A few differences were noted between the sediment collected in 2008 and the
sediment collected in 2019. The sediment was less dense in 2019 and the concentration of organic
phosphorus (organic-P) was higher in 2019. This is perhaps an outcome of the dense curly-leaf pondweed
growth and die-off in recent years. Curlyleaf growth and die-off could lead to a build-up of a more
organic sediment with higher organically-bound phosphorus. A summary of the sediment chemistry is
provided in Table 1 below.
Table 1 Average phosphorus concentrations in sediment cores collected in Lake Cornelia
in 2008 and in 2019
Units are mg phosphorus per gram of dry sediment.
Basin
Average-Top 6 cm
Mobile P
(mg/g dry)
Aluminum-P
(mg/g dry)
Organic-P
(mg/g dry)
Calcium-P
(mg/g dry)
2008 2019 2008 2019 2008 2019 2008 2019
North Cornelia 0.29 0.23 0.09 0.05 0.18 0.71 0.25 0.21
South Cornelia 0.13 0.05 0.07 0.05 0.19 0.72 0.23 0.22
Service Layer Credits:
!(
!(
!(
!(
!(
#*
#*
#*
#*
#*
Corn1
Corn2
Corn3
Corn4
Corn6
sc1
sc2
sc3
sc4
sc5
W 66th St
Point DrSouthdale RdRoycar Rd
W 65th St
Hillcrest LaWooddale AveVal
leyv
iew
Rd
W
e
s
t
S
h
o
r
e
D
r
Oaklawn Ave
W 68th St
Creston R
d
Southcrest DrBulfanz Rd Va
l
l
e
y
V
i
ew
Rd
Laguna Dr
Cornelia DrW
6
4
t
h
S
t
Barr Footer: ArcGIS 10.6.1, 2019-07-11 09:20 File: I:\Client\Nine_Mile_Creek_WD\Work_Orders\23270634_Project\Maps\Misc_2019\Lake_Cornelia\Fig01 Sediment Coring Locations.mxd User: MJM3
SEDIMENT CORING LOCATIONSLake CorneliaEdina, Minnesota
250 0 250
Feet
!;N
!(Sediment Core - 2008
#*Sediment Core - 2019
Approximate Water Depth
0 - 1 Feet
1 - 2 Feet
2 - 3 Feet
3 - 4 Feet
4 - 5 Feet
5 - 6 Feet
6 - 7 Feet
7 - 8 Feet
Figure 1Imagery Source: NearMap, April 2019
6
4 Alum Application and Dosing Plan
Collecting sediment cores and conducting phosphorus fractionation of the lake-specific sediments is
important in developing a targeted application and dosing plan that will maximize effectiveness and
longevity of an alum treatment. One of the purposes of sediment coring and phosphorus fractionation is
the calculation of an alum dose that can bind the mobile phosphorus fraction and prevent internal
loading (the mobile phosphorus fraction is the fraction that releases during low oxygen conditions).
Another purpose is to understand and plan for the fraction of organically- bound phosphorus in the
sediments, as that fraction decays over time (e.g., 10% a year) and becomes part of the mobile P pool.
Sediment analysis results indicate the concentration of organically-bound phosphorus is high in North
and South Lake Cornelia and hence a split alum (aluminum) application is recommended. Half of the alum
(aluminum) would be applied followed by another treatment 5 years after the initial treatment. This would
allow the organic-P to decay into mobile phosphorus and then the second application would bind the
decayed/new mobile phosphorus. It should be noted that alum (aluminum) cannot directly bind and
immobilize the organically-bound aluminum. Finally, it should also be noted that mobile P was lower in
2019. This is likely due to the timing of sediment collection, cores were collected in the first week in June
in 2019 (some phosphorus likely had already released from the sediment thereby reducing the
concentration of phosphorus in the sediment) while the cores from 2008 were collected the last week of
October (phosphorus should have settled back down into the sediments). Sediment chemistry data from
both 2008 and 2019 were used to calculate alum doses.
Using the mobile-P data collected from North and South Lake Cornelia and the methods of Pilgrim et al.,
2007, alum doses were calculated to achieve an approximately 85 percent internal phosphorus loading
reduction (Table 2). This target was used for the 2018 UAA update, as internal load reduction targets
greater than 85 percent often lead to significantly higher doses and costs and the additional benefit is
negligible and not cost efficient. The total alum dose was calculated to immobilize mobile phosphorus in
the top six centimeters of sediment. Because Lake Cornelia is shallow, it will be necessary to apply both
alum and sodium aluminate to buffer pH in a near neutral (pH 7.0) range that is safe for fish and aquatic
life. The application area includes all areas within both basins with water depths greater than two feet.
The recommended dose for 2019 for North Cornelia is 55 grams of aluminum per square meter of lake-
bottom and 22 grams of aluminum per square meter of lake-bottom for South Cornelia. For comparison
purposes, the alum dose for the 2019 alum treatment at Normandale Lake was approximately 30 grams
per square meter of lake bottom. For the recommended dose in 2019, a comparison to Normandale Lake
shows that a higher dose is being applied to North Cornelia while a slightly lower dose is being applied to
South Cornelia (see Table 2). These differences are based upon the concentration of mobile-P in the
sediments of these lakes. The collection of sediment cores is recommended one year prior to the
application of the second dose to North and South Cornelia to confirm that the second application is
warranted and to update the dose, as necessary, based on mobile-P concentrations.
7
Table 2 Recommended alum and sodium aluminate application dose for 2019 and the
total application dose after a second application
Bay Treatment
Area (ac)
Mobile-P
(g m-2 cm-1)1
Aluminum
Dose (g m-2)
Gallons per Acre Total Gallons
Alum Sodium
Aluminate Alum Sodium
Aluminate
Recommended Dose for 2019
North Cornelia 15.6 0.33 55 439 220 6,850 3,425
South Cornelia 27.7 0.13 22 174 87 4,817 2,409
Total 11,667 5,834
Full Dose: Amount of Alum Application Applied in Total After Second Application2
North Cornelia 15.6 0.33 110 878 439 13,700 6,850
South Cornelia 27.7 0.13 44 348 174 9,634 4,817
Total 23,335 11,667
1 Mobile phosphorus concentration consists of the iron-bound and loosely-sorbed phosphorus concentration. Value is
the average of the top 6 cm of sediment which was used for dosing purposes.
2 Total application dose assumes the second application will be at a dosing rate consistent with the recommended 2019
dose. Additional sediment core collection and analysis is recommended prior to second application to confirm dosing
rate.
8
5 Cost Estimate
5.1 Opinion of Probable Cost
A planning-level opinion of probable cost was developed for the alum treatment based on previous
project applications and communication with an alum application contractor. The cost estimate includes
individual and total costs for a split dose in 2019 and 2024, in 2019 dollars. The opinion of cost is intended
to provide assistance in planning and budgeting, but should not be assumed as an absolute value. The
estimated costs are summarized in Table 3.
Table 3 Engineer’s opinion of probable cost
Basin Treatment Cost1
($)
Treatment Cost Range2($) Estimated Life
of Project
Low High
2019 Alum Treatment- 1st Dose $116,000 $104,000 $139,000 5 years
2024 Alum Treatment- 2nd Dose $116,000 $104,000 $139,000 5 years
Total (combination of two
treatments) $232,000 $208,000 $278,000 10 years
1 Treatment cost includes mobilization cost estimate ($15,300), alum application cost estimate ($69,700),
engineering costs for bidding, contracting, pH monitoring and contractor oversight ($20,000), and 10% overall
contingency.
2 Opinion of probable cost is considered a Class 2 cost estimate corresponding to standards established by
the Association for the Advancement of Cost Engineering (AACE), with a range between -10% and +20%.
5.2 Cost-Benefit Analysis
A cost-benefit analysis of potential management activities in Lake Cornelia and its tributary watershed was
conducted as part of the recently completed Use Attainability Analysis (UAA) for Lake Cornelia (updated
from 2010) and Lake Edina (first version). Results of the analysis showed that the cost-benefit ratio for the
alum treatment is low in comparison with other watershed management practices, indicating a high
benefit per dollar spent. Published literature supports the conclusion that alum treatment costs are often
low compared to the benefit that is received in terms of phosphorus reductions (see Bartodziej et al.,
2017). As part of this analysis, the cost-benefit ratio was recomputed based on the estimated alum
treatment costs identified in Section 5.1. The methodology and results are summarized below.
Estimated costs for the alum treatment were annualized to help compare the cost-benefit ratio. The
annualized cost for each management alternative is based on anticipated maintenance, replacement
costs, and anticipated useful life-span of the projects/treatments. A 3% interest rate was assumed. The
annualized cost was calculated as the value of ‘n’ equal, annual payments, where ‘n’ is the anticipated
useful life-span of the project or treatment.
9
For the cost-benefit analysis, two approaches were considered to quantify the benefits of each of the
evaluated management activities. The first approach quantifies the benefit in terms of phosphorus
removed (in pounds) during the time period of April through September (i.e., phosphorus that did not
enter the lake system as a result of the management practice). Table 4 summarizes the annualized costs
per pound of phosphorus removed in North and South Cornelia. The second approach quantifies the
benefit in terms of reduced summer average total phosphorus concentration in North and South Cornelia
(June through September). Table 5 summarizes the annualized costs per unit reduction in in-lake summer
average total phosphorus concentration (µg/L) from conducting an alum treatment.
Table 4 Summary of annualized cost per pound of total phosphorus removed from
conducting an alum treatment in Lake Cornelia
Basin
Estimated Treatment
Cost1 ($)
Annualized
Estimated Treatment
Cost ($/year)
Current TP Load2
(lb)
Average TP Load
Removed (lb)2,3
Annualized Cost
per Pound of TP
Removed
($/lb)
Low High Low High Low High Low High Low6 High6
North
Cornelia $61,000 $82,000 $13,745 $18,370 456 673 54 222 $340 $62
South
Cornelia4 $43,000 $57,000 $9,665 $12,918 442 525 86 169 $150 $57
Both $104,000 $139,000 $23,411 $31,289 898 1,198 140 391 $223 $60
1 Estimated cost range reflects -10% to +20% of point cost estimate for each lake.
2 Range from the 2018 Lake Cornelia UAA for modeling years 2015, 2016, and 2017. Load for modeling period early-
to mid-April through September 30.
3 Load reduction for modeling period early to mid-April through September 30.
4 Part of the load reduction in South Cornelia is a result of the load reduction in North Cornelia.
5 Cost-benefit calculation assumes that the longevity of the alum treatment is 5 years.
6 High estimate assumes the low annual phosphorus load reduction and the high cost and the low estimate assumes
high annual phosphorus load reduction and the low cost.
Table 5 Summary of annualized cost per unit reduction (µg/L) in summer-average total
phosphorus concentration.
Basin
Treatment Cost1 ($)
Annualized
Estimated Treatment
Cost ($/year)
Current In-Lake
Summer-Average
Phosphorus
Concentration (µg/L)
In-Lake
Phosphorus
(µg/L)
Reduction2
Cost Benefit
($/µg/L P
Reduction)3
Low High Low High Low High Low High Low4 High4
North
Cornelia $61,000 $82,000 $13,745 $18,370 126 185 21 42 $875 $327
South
Cornelia4 $43,000 $57,000 $9,665 $12,918 118 178 25 85 $517 $114
1 Estimated cost range reflects -10% to +20% of point cost estimate.
2 From the 2018 Lake Cornelia UAA for modeling years 2015, 2016, and 2017.
3 Cost-benefit calculation assumes that the longevity of the alum treatment is 5 years.
4 High estimate assumes the low annual phosphorus benefit and the high cost and the low estimate assumes high
annual phosphorus benefit and the low cost.
10
6 Site Access
Site access options were evaluated as part of this study, which included review of land use and
topographic conditions adjacent to the lake and communications with City of Edina staff and an alum
application contractor. Direct access to the lake for barge access, rather than using a crane to place the
treatment barge on the water, is preferable and would likely be more cost effective. A staging area near
the lake where temporary tanks can be placed to hold alum and sodium aluminate will be necessary.
These tanks are used to refill the alum treatment barge. Potential site access locations are shown in
Figure 2, based on these considerations. A city-owned parcel at the intersection of Laguna Drive and
Wooddale Avenue has been identified as a potential barge launching and staging area for treatment of
South Cornelia. The area just north or south of the fishing pier in Rosland Park has been identified as a
potential barge launching and staging area for treatment of North Cornelia. The fishing pier would be
useful for barge refilling as it would allow the barge to remain in the deeper part of the lake during
refilling and hence there would be reduced sediment disturbance. Both of these sites have been discussed
with the City of Edina as potential boat launching and staging areas with use by the contractor contingent
upon written approval by the City of Edina Parks Department.
Service Layer Credits:
Potential Staging Area and Access Pointfor Treatment Barge
Potential Staging Areaand Access Pointfor Treatment Barge
W 66th St
Point DrSouthdale RdRoycar Rd
W 65th St
Hillcrest LaWooddale AveVal
leyv
iew
Rd
W
e
s
t
S
h
o
r
e
D
r
Oaklawn Ave
W 68th St
Creston Rd
Southcrest DrBulfanz Rd Va
l
l
e
y
V
i
ew
Rd
Laguna Dr
Cornelia DrW
6
4
t
h
S
t
Barr Footer: ArcGIS 10.6.1, 2019-07-11 09:21 File: I:\Client\Nine_Mile_Creek_WD\Work_Orders\23270634_Project\Maps\Misc_2019\Lake_Cornelia\Fig02 Potential Staging Locations.mxd User: MJM3
POTENTIAL STAGING LOCATIONSLake CorneliaEdina, Minnesota
250 0 250
Feet
!;N
Parcel Boundaries
Approximate Water Depth
0 - 1 Feet
1 - 2 Feet
2 - 3 Feet
3 - 4 Feet
4 - 5 Feet
5 - 6 Feet
6 - 7 Feet
7 - 8 Feet
Figure 2Imagery Source: NearMap, April 2019
12
7 Permitting
With the exception of a written letter of approval by the City of Edina Parks Director for site access, no
additional local permitting requirements have been identified for the alum application. The Minnesota
Pollution Control Agency (MPCA) typically requires submittal of a notification letter regarding the planned
alum treatment. This letter typically includes a basis of the need for the alum treatment, dosing, and
application timing. The MPCA provides a written notification to proceed.
13
8 Timing
The effectiveness of an alum treatment can be influenced by the timing of application. Typically, alum
treatments are conducted in the spring or fall to avoid alum floc getting caught up on vegetation and to
avoid floc settling impedance during an algal bloom (the floc can get stuck to buoyant algae). The alum
treatment in Lake Cornelia is tentatively planned for October 2019. During the fall, the primary limitations
for treatment are ice formation and water temperature (temperature needs to be approximately
42 degrees Fahrenheit or greater).
14
References
Barr Engineering Co. 2019. Lake Cornelia and Lake Edina Water Quality Study Use Attainability Analyses
for Lake Cornelia (updated from 2010) and Lake Edina (first version).
Bartodziej, B., Blood, S., and K. Pilgrim. 2017. Aquatic plant harvesting: An economical phosphorus
removal tool in an urban shallow lake. J. Aquat. Plant Manage. 55: 26-34.
Natarajan, P. and J. Gulliver. 2019. Assessment of Internal Phosphorus Loading in Swimming Pool Pond
and Point of France Pond, City of Edina. St. Anthony Falls Laboratory Project Report No. 587.
Pilgrim, K., Huser, B., and P. Brezonik. 2007. A method for comparative evaluation of whole-lake and
inflow alum treatment. Wat. Research, 41, 1215-1224.