Pocatello's Halliday Street Drain Experience
Calibrating storm drain flows for pollutant discharge load calculations for NPDES compliance
Halliday
is a major drain that collects stormwater from Pocatello’s urban area and
discharges to the Portneuf River via an 84-inch pipe. Construction of the
Halliday storm drain was completed prior to the National Pollutant Discharge
Elimination System (NPDES) permitting requirements for monitoring pollutant
loads from stormwater runoff to the river. As built, the drain had no monitoring
equipment whatsoever. Pocatello city staff have installed monitoring equipment
and conducted a coordinated program to calibrate the flow gaging
equipment.
Background
The
Pocatello Urbanized Area (PUA) received its initial NPDES Phase II municipal
separate storm sewer system (MS4) permit in December 2006. The PUA permit was
the first Phase II MS4 permit issued in Idaho and only the second Phase II
permit issued by the EPA Region 10 in Idaho or Alaska. The permit was issued by
Region 10 (which includes Idaho, Oregon, Washington, and Alaska) because the
state of Idaho has not sought NPDES primacy.
The
2006 PUA MS4 permit contains requirements to monitor stormwater quantity and
quality to aid in the selection and evaluation of best management practices
(BMPs). This article describes the water-quantity and -quality monitoring
approach being implemented to satisfy the monitoring requirements contained in
the permit.
The
EPA promulgated NPDES stormwater regulations in 1973, 1976, 1979, and 1984.
These regulations resulted in extensive litigation, but little action or
improvement, in stormwater discharges to the nation’s waters. Requirements for a
comprehensive stormwater management program were included in the 1987 amendments
to the federal Clean Water Act, section 402(p).
Section
402(p) established a phased and tiered approach to permitting stormwater
discharges from industrial, municipal, and other sources. Phase I of the
stormwater program regulates industrial and large and medium municipal
dischargers, and Phase II covers all discharges not included in Phase I. Both
Phase I and II require MS4s to develop, implement, and enforce a stormwater
management program that will reduce the discharge of pollutants to the “maximum
extent practicable” and that “satisfy the water quality requirements of the
Clean Water Act (40CFR122.44) (a).”
The
small, or Phase II, MS4 permittees must address six minimum measures, including
public education and outreach, public involvement/participation, illicit
discharge detection and elimination, construction-site runoff controls,
post-construction stormwater management for developed and redeveloped areas, and
pollution prevention/good housekeeping for municipal operations. Additionally,
small MS4s are required to identify BMPs needed to implement the six minimum
measures and “measurable goals” to demonstrate effectiveness (40CFR122.34(d)),
and additional measures are based on an approved total maximum daily load (TMDL)
or similar analysis needed to protect water quality (40CFR122.34(e)(1)). The
water-quality and -quantity monitoring requirements in the permit are present to
help us develop and implement BMPs that protect water
quality.
Pocatello
Urbanized Area
PUA
boundaries incorporate portions of the cities of Pocatello and Chubbuck, part of
Bannock County, and a portion of the southeastern district of the Idaho
Transportation Department District in southeastern Idaho. Boundaries for NPDES
Phase II permittees are based on populations of more than 50,000, but fewer than
the 100,000-population designation for large MS4s.
Mapping
of the stormwater system, including global positioning system (GPS)
documentation of all aspects of the system, has been initiated. Nearly 200
stormwater outfalls into jurisdictional waters of the US are indicated on old
city maps, necessitating an ongoing and concerted effort to locate and map these
features. In addition, all catchments and stormwater sewer manholes are being
located by GPS and incorporated into the city’s geographic information system
layers. This process will allow us to define each stormwater basin or watershed
for subsequent analyses of BMPs based on outfall pollutant loads. Annual
waste-load management for all of these outfalls is an objective, if not an
outright requirement, of the current stormwater permit. Annual waste-load
management cannot be achieved without knowledge of the locations of all of the
stormwater outfalls, potential contributing sources in each watershed, and flow
levels and contaminant loads for every outfall. This information will ultimately
be used by PUA stormwater managers to select appropriate nonstructural or
structural BMPs to meet Idaho water-quality standards for the Portneuf
River.
Monitoring
Requirements in the Permit
Included
in the federal permit mandates are stormwater monitoring requirements for four
designated stormwater outfalls in the city of Pocatello: Halliday Street, Lander
Street, Day Street, and Pocatello Creek. The Halliday Street storm drain is the
largest of the city’s approximately 200 storm outfalls. The Halliday Street
storm drain, installed in 2002, consists of 10,500 linear feet of concrete pipe
ranging in size from 36 to 84 inches. It drains more than 3 square miles of
Pocatello, including significant portions of the Idaho State University campus.
Additional monitoring is required in the Portneuf River downstream of each of
the four designated outfalls. In the fall of 2007, city staff developed a
sampling protocol for the four designated stormwater outfalls and the river
below each of them. Because the Portneuf River through this part of Pocatello is
conveyed in a concrete channel with 20-foot-high walls and 8-foot-high chain
link fences, sampling points take advantage of city street bridges for access to
the channel and the river.
Provisions
of the NPDES permit require the city to determine annual pollutant loading for
each pollutant of concern and an overall estimate of the amount of each
pollutant discharged to the Portneuf River from the entire PUA stormwater
system.
Pollutants of concern for the PUA are identified in the permit and in the
Portneuf River TMDL documentation as sediment, nitrogen, phosphorus, oil and
grease, and E.
coli
bacteria. To estimate these loads, flow volumes during storm events must be
correlated with the analytical results of the pollutants of concern. The
Halliday Street storm drain was selected as the first location to be
instrumented for stormwater monitoring and flow recording.
Equipment
The
Halliday stormwater monitoring station features data acquisition of flow, an
automatic sampler, and telemetry to connect these devices to the Internet, which
allows them to be monitored and controlled remotely. This system allows for
real-time monitoring of water levels in the Halliday outfall box and for
automatic remote notification of defined events. A schematic of the electronic
components of the monitoring system is depicted in Figure 1. Components are
described
below.
Level
Sensor.
We selected a Siemens Milltronics ultrasonic level sensor (EnviroRanger ERS 500)
to measure stage level in the vault near the outfall of the Halliday storm
drain. This level sensor has a downward-looking transducer that is mounted at
the top of the concrete vault at the end of the 84-inch pipe. Water from the
84-inch pipe flows into the 6-foot by 10-foot vault and back into a 6-foot
section of 8-inch pipe before dropping into the concrete channel (a vertical
distance of a few feet, depending on river flow levels). Of the many options for
measuring level in the storm drain, noncontacting ultrasonic technology was the
preferred choice because it simplified mounting issues in the vault and was
immune to problems such as drift and fouling. This type of sensor has an
acceptable resolution (0.08 inch) and accuracy (0.24 inch) for the application,
and they have proven to be reliable for the city of Pocatello, which uses the
EnviroRanger throughout its network of several dozen wastewater lift
stations.
Data
Logger.
The level sensor produces a 4-20 mA data output that is monitored with a CR800
measurement and control data logger from Campbell Scientific Inc. The data
logger is powered by a battery-backed 12-volt DC supply and has the capacity to
store 4 megabytes of information. The data logger is programmed to read the
output from the level sensor every 10 seconds and can convert the 4-20 mA
signals to inches and cubic feet per second based on calibration equations
maintained on the remote server. The CR800 then logs the average level and flow
at 10-minute intervals. The data logger also monitors conditions within the
equipment enclosure, such as air temperature, grid power status, and battery
voltage. The data logger has a switched control output that is connected to the
automatic sampler, thereby allowing samples to be collected based on flow
conditions or a trigger controlled via telemetry.
 |
| Figure 2. Knaack toolbox that houses all field monitoring equipment |
Automatic
Sampler.
Water samples are collected from the outfall using an Isco 6700 pump sampler
housed within a Knaack lockable toolbox for security (Figure 2). The sampler
features a peristaltic pump that draws a liquid sample from the storm drain
through polyethylene tubing. This sampler provides a sample velocity in the
tubing that exceeds the EPA-recommended speed of 2 feet per second. Samples are
delivered sequentially to 1-liter glass bottles based on a trigger received from
the data logger. An electric heater within this toolbox maintains temperature
above freezing during winter.
Telemetry.
An industrial-grade digital cellular data radio modem (Airlink EDGE Raven)
provides communication access to the CR800 dataloger over the Internet. This
full-duplex telephone modem transmits data to the local cellular tower using
either a GPRS (General Packet Radio Service) or EDGE (Enhanced Data rates for
GSM Evolution) network. For our application, the Raven has been set up with a
static IP address for direct connection to the Internet. The Raven operates off
of 12-volt DC, and due to its relatively low power demands (20 mA dormant, 130
mA transmit/receive), it is left powered up at all times. The Raven is accessed
through a contract with a cellular service provider, which costs about $20 per
month for 5 megabytes of data use.
Monitoring
Server at the Office.
A Microsoft Windows-based computer running the data logger support software,
Loggernet from Campbell Scientific Inc., provides the communication link with
the data logger over the Internet. Loggernet is programmed to call the station
on an hourly basis and collect data acquired since the last connection. Data are
saved to the hard drive of the monitoring server. Loggernet also performs
regularly scheduled maintenance tasks, such as checking and adjusting the data
logger’s internal clock. The logged data are managed on the server using Vista
Data Vision, a software tool that imports the recently acquired data to a MySQL
database, and allows for access by means of a Web browser to the data over the
Internet. The user can specify trend plots and data tables for any time period
using Vista Data Vision. The software continuously monitors the status of
specified fields in the database and is programmed to send e-mail and text
messages when flow values exceed specified set points. This provides a means to
notify field personnel when runoff from a storm exceeds a particular level
(e.g., 4 inches and 8 inches). We set up the system to notify users via e-mail
or cell phone when water levels have crossed these trigger
points.
The
alarm and notification features are important because supplemental sampling for
oil and grease and for E.
coli
are required at the Halliday site and its accompanying down river sampling
location, as well as at other locations throughout the storm sewer system and
their companion river sampling locations. The alarm also functions as a general
stormwater event monitor, allowing for notification of city staff during storm
events and facilitating sampling of other stormwater outfalls as required by the
permit. Sampling is conducted by city staff at all the required stormwater and
river monitoring stations as long as stormwater flow is
continuous.
 |
| Figure 3. Water discharge from the Halliday storm drain to the Portneuf River |
Approach
As
part of the process to determine total annual discharge and develop an estimate
of the pollutant load for each of the pollutants of concern, the city selected
the Halliday Street storm drain for our initial sampling effort. Halliday is the
largest stormwater pipe, has easy access at the end of a city street, and has a
readily accessible electric power supply. The Halliday Street drain is unique in
its large size but otherwise is a representative stormwater outfall. The
Halliday Street drain has a steep slope in the upper reaches and a slope of
approximately 0.001 in the last 800 feet. The total length of the system is over
10,500 feet from the east side of Pocatello to its discharge point into the
Portneuf River in the concrete channel. Twelve feet from the discharge point,
the 84-inch pipe is discharged into a 6-foot by 10-foot grated vault. Flows from
the vault enter a 6-foot section of 84-inch pipe for discharge into the channel
(Figure 3). PUA permit conditions allow for either automatic sampling or grab
samples for stormwater for pollutant load characterization. We elected to
instrument the Halliday Street storm drain so that it would function as a storm
event alarm for flows throughout the entire system.
To
calibrate the Halliday stormwater outfall and allow for calculation of total
flow and total pollutant load, city staff developed a calibration flow protocol
using an offline city well. City Water Department staff estimated that, by using
the three available booster pumps, water could be routed from a 3 million gallon
water-storage tank and pump a total of 6,000 gallons per minute (gpm) through
the storm drain. To enable the city to accommodate the anticipated flows and
pressures during the calibration event, Rain for Rent (Idaho Falls, ID) was
contracted to provide high-density polyethylene (HDPE) pipe and fittings. HDPE
was selected because of its flexibility, a pressure rating of 160 pounds per
square inch (psi), corrosion resistance, zero leakage, extreme durability, and
resistance to weather, chemicals, fatigue, and surges.
The
booster pump was isolated from the rest of the municipal potable water supply by
valving so all the water measured at the booster was directed to the Halliday
stormwater drain. During the morning of the test, it was fairly cold, and there
had been no precipitation during the previous 24 hours, ensuring that the
measurements of flow were not compromised by outside flow influences.
 |
| Figure 4. HDPE Pipe assembly at the Halliday test site |
 |
| Figure 5. HDPE field assembly and tie down |
The
HDPE was premeasured and fused by Rain for Rent at Idaho Falls for assembly on
the job site (Figure 4). The pipe was heat fused with the Tracstar 500 Fusion
Machine. Each of the six flange adapters was rated to 150 psi and individually
fused to the lengths of pipe. Each fuse took approximately 30 minutes. At 6,000
gpm, the water was experiencing only 2.5 psi of friction loss through the 93.5
feet of 12-inch HDPE pipe. Water velocity was at 19.5 feet per second.
Increasing amounts of water were pumped to the storm drain until the 6,000-gpm
capacity was reached. Measurements of flow at the pump and in the vault were
taken as flow increased and as flows were subsequently decreased. The pipe was
held in place by several cubic yards of road base (Figure
5).
Results
and Discussion
We
used Manning’s equation for flow in open pipe to correlate estimated flow rates
from the booster pumps with the depth of water flowing in the pipe as measured
by the Milltronics equipment. Assuming a Manning’s n
of 0.010 and the known pipe slope of 0.001 the results were as shown in Table
1.
As
is evident from the data, the calculated flow rate is approximately 10% below
the flow rate recorded at the booster pump. Several variables may account for
all or a portion of this discrepancy:
-
The accuracy of the flow meter at the pump station is unknown, leading to
potential errors in measuring actual flows.
- 6-foot by 10-foot vault near the outfall end of the Halliday Street storm
sewer complicated accurate measurement of calibration flows. This change in
conveyance configuration shape and slope may have complicated accurate
measurements of depth of flow. Our depth measurements with the Milltronics also
varied due to wave action in the vault and/or the stability of the sensor.
Graphic presentation of the various data components is provided in Figure 6. We
calculated an average depth to use in the equation.
- We assumed a Manning’s n
of 0.010 for smooth concrete pipe.
Even
with all the variables in play, the comparison of the data showed the flow rate
tracking parallel to the curve predicted by the Manning equation. We intend to
use the Manning equation with an n factor of 0.010 to calculate our flow rates
and total flow during storm events, because we believe that as the depth of flow
in the pipe increases at higher flow rates, the effect of the variables noted
above will diminish. During the calibration test, we noticed that the Miltronics
level transducer, which was suspended from the grating with a cable, was swaying
somewhat with the wind. We subsequently changed the transducer mounting in the
vault to a solid metal arm, which appears to have reduced the variability in
level readings.
The
monitoring installation at the Halliday storm drain will allow us to calculate
total flow during individual storm events and subsequently calculate the total
load for the constituents of concern (sediment, phosphorus, nitrogen, oil and
grease, and bacteria) using an integral of the area under the flow meter curve.
The calibration event undertaken by the city bolsters our confidence in the
accuracy of the flow data used in our stormwater management program.
Author's Bio: John W. Sigler, Ph.D., CEP, QEP, is senior environmental coordinator for the city of Pocatello, ID.
October 2008
Pocatello's Halliday Street Drain Experience
Calibrating storm drain flows for pollutant discharge load calculations for NPDES compliance
Halliday
is a major drain that collects stormwater from Pocatello’s urban area and
discharges to the Portneuf River via an 84-inch pipe. Construction of the
Halliday storm drain was completed prior to the National Pollutant Discharge
Elimination System (NPDES) permitting requirements for monitoring pollutant
loads from stormwater runoff to the river. As built, the drain had no monitoring
equipment whatsoever. Pocatello city staff have installed monitoring equipment
and conducted a coordinated program to calibrate the flow gaging
equipment.
Background
The
Pocatello Urbanized Area (PUA) received its initial NPDES Phase II municipal
separate storm sewer system (MS4) permit in December 2006. The PUA permit was
the first Phase II MS4 permit issued in Idaho and only the second Phase II
permit issued by the EPA Region 10 in Idaho or Alaska. The permit was issued by
Region 10 (which includes Idaho, Oregon, Washington, and Alaska) because the
state of Idaho has not sought NPDES primacy.
The
2006 PUA MS4 permit contains requirements to monitor stormwater quantity and
quality to aid in the selection and evaluation of best management practices
(BMPs). This article describes the water-quantity and -quality monitoring
approach being implemented to satisfy the monitoring requirements contained in
the permit.
The
EPA promulgated NPDES stormwater regulations in 1973, 1976, 1979, and 1984.
These regulations resulted in extensive litigation, but little action or
improvement, in stormwater discharges to the nation’s waters. Requirements for a
comprehensive stormwater management program were included in the 1987 amendments
to the federal Clean Water Act, section 402(p).
Section
402(p) established a phased and tiered approach to permitting stormwater
discharges from industrial, municipal, and other sources. Phase I of the
stormwater program regulates industrial and large and medium municipal
dischargers, and Phase II covers all discharges not included in Phase I. Both
Phase I and II require MS4s to develop, implement, and enforce a stormwater
management program that will reduce the discharge of pollutants to the “maximum
extent practicable” and that “satisfy the water quality requirements of the
Clean Water Act (40CFR122.44) (a).”
The
small, or Phase II, MS4 permittees must address six minimum measures, including
public education and outreach, public involvement/participation, illicit
discharge detection and elimination, construction-site runoff controls,
post-construction stormwater management for developed and redeveloped areas, and
pollution prevention/good housekeeping for municipal operations. Additionally,
small MS4s are required to identify BMPs needed to implement the six minimum
measures and “measurable goals” to demonstrate effectiveness (40CFR122.34(d)),
and additional measures are based on an approved total maximum daily load (TMDL)
or similar analysis needed to protect water quality (40CFR122.34(e)(1)). The
water-quality and -quantity monitoring requirements in the permit are present to
help us develop and implement BMPs that protect water
quality.
Pocatello
Urbanized Area
PUA
boundaries incorporate portions of the cities of Pocatello and Chubbuck, part of
Bannock County, and a portion of the southeastern district of the Idaho
Transportation Department District in southeastern Idaho. Boundaries for NPDES
Phase II permittees are based on populations of more than 50,000, but fewer than
the 100,000-population designation for large MS4s.
Mapping
of the stormwater system, including global positioning system (GPS)
documentation of all aspects of the system, has been initiated. Nearly 200
stormwater outfalls into jurisdictional waters of the US are indicated on old
city maps, necessitating an ongoing and concerted effort to locate and map these
features. In addition, all catchments and stormwater sewer manholes are being
located by GPS and incorporated into the city’s geographic information system
layers. This process will allow us to define each stormwater basin or watershed
for subsequent analyses of BMPs based on outfall pollutant loads. Annual
waste-load management for all of these outfalls is an objective, if not an
outright requirement, of the current stormwater permit. Annual waste-load
management cannot be achieved without knowledge of the locations of all of the
stormwater outfalls, potential contributing sources in each watershed, and flow
levels and contaminant loads for every outfall. This information will ultimately
be used by PUA stormwater managers to select appropriate nonstructural or
structural BMPs to meet Idaho water-quality standards for the Portneuf
River.
Monitoring
Requirements in the Permit
Included
in the federal permit mandates are stormwater monitoring requirements for four
designated stormwater outfalls in the city of Pocatello: Halliday Street, Lander
Street, Day Street, and Pocatello Creek. The Halliday Street storm drain is the
largest of the city’s approximately 200 storm outfalls. The Halliday Street
storm drain, installed in 2002, consists of 10,500 linear feet of concrete pipe
ranging in size from 36 to 84 inches. It drains more than 3 square miles of
Pocatello, including significant portions of the Idaho State University campus.
Additional monitoring is required in the Portneuf River downstream of each of
the four designated outfalls. In the fall of 2007, city staff developed a
sampling protocol for the four designated stormwater outfalls and the river
below each of them. Because the Portneuf River through this part of Pocatello is
conveyed in a concrete channel with 20-foot-high walls and 8-foot-high chain
link fences, sampling points take advantage of city street bridges for access to
the channel and the river.
Provisions
of the NPDES permit require the city to determine annual pollutant loading for
each pollutant of concern and an overall estimate of the amount of each
pollutant discharged to the Portneuf River from the entire PUA stormwater
system.
Pollutants of concern for the PUA are identified in the permit and in the
Portneuf River TMDL documentation as sediment, nitrogen, phosphorus, oil and
grease, and E.
coli
bacteria. To estimate these loads, flow volumes during storm events must be
correlated with the analytical results of the pollutants of concern. The
Halliday Street storm drain was selected as the first location to be
instrumented for stormwater monitoring and flow recording.
Equipment
The
Halliday stormwater monitoring station features data acquisition of flow, an
automatic sampler, and telemetry to connect these devices to the Internet, which
allows them to be monitored and controlled remotely. This system allows for
real-time monitoring of water levels in the Halliday outfall box and for
automatic remote notification of defined events. A schematic of the electronic
components of the monitoring system is depicted in Figure 1. Components are
described
below.
Level
Sensor.
We selected a Siemens Milltronics ultrasonic level sensor (EnviroRanger ERS 500)
to measure stage level in the vault near the outfall of the Halliday storm
drain. This level sensor has a downward-looking transducer that is mounted at
the top of the concrete vault at the end of the 84-inch pipe. Water from the
84-inch pipe flows into the 6-foot by 10-foot vault and back into a 6-foot
section of 8-inch pipe before dropping into the concrete channel (a vertical
distance of a few feet, depending on river flow levels). Of the many options for
measuring level in the storm drain, noncontacting ultrasonic technology was the
preferred choice because it simplified mounting issues in the vault and was
immune to problems such as drift and fouling. This type of sensor has an
acceptable resolution (0.08 inch) and accuracy (0.24 inch) for the application,
and they have proven to be reliable for the city of Pocatello, which uses the
EnviroRanger throughout its network of several dozen wastewater lift
stations.
Data
Logger.
The level sensor produces a 4-20 mA data output that is monitored with a CR800
measurement and control data logger from Campbell Scientific Inc. The data
logger is powered by a battery-backed 12-volt DC supply and has the capacity to
store 4 megabytes of information. The data logger is programmed to read the
output from the level sensor every 10 seconds and can convert the 4-20 mA
signals to inches and cubic feet per second based on calibration equations
maintained on the remote server. The CR800 then logs the average level and flow
at 10-minute intervals. The data logger also monitors conditions within the
equipment enclosure, such as air temperature, grid power status, and battery
voltage. The data logger has a switched control output that is connected to the
automatic sampler, thereby allowing samples to be collected based on flow
conditions or a trigger controlled via telemetry.
|
| Figure 2. Knaack toolbox that houses all field monitoring equipment |
Automatic
Sampler.
Water samples are collected from the outfall using an Isco 6700 pump sampler
housed within a Knaack lockable toolbox for security (Figure 2). The sampler
features a peristaltic pump that draws a liquid sample from the storm drain
through polyethylene tubing. This sampler provides a sample velocity in the
tubing that exceeds the EPA-recommended speed of 2 feet per second. Samples are
delivered sequentially to 1-liter glass bottles based on a trigger received from
the data logger. An electric heater within this toolbox maintains temperature
above freezing during winter.
Telemetry.
An industrial-grade digital cellular data radio modem (Airlink EDGE Raven)
provides communication access to the CR800 dataloger over the Internet. This
full-duplex telephone modem transmits data to the local cellular tower using
either a GPRS (General Packet Radio Service) or EDGE (Enhanced Data rates for
GSM Evolution) network. For our application, the Raven has been set up with a
static IP address for direct connection to the Internet. The Raven operates off
of 12-volt DC, and due to its relatively low power demands (20 mA dormant, 130
mA transmit/receive), it is left powered up at all times. The Raven is accessed
through a contract with a cellular service provider, which costs about $20 per
month for 5 megabytes of data use.
Monitoring
Server at the Office.
A Microsoft Windows-based computer running the data logger support software,
Loggernet from Campbell Scientific Inc., provides the communication link with
the data logger over the Internet. Loggernet is programmed to call the station
on an hourly basis and collect data acquired since the last connection. Data are
saved to the hard drive of the monitoring server. Loggernet also performs
regularly scheduled maintenance tasks, such as checking and adjusting the data
logger’s internal clock. The logged data are managed on the server using Vista
Data Vision, a software tool that imports the recently acquired data to a MySQL
database, and allows for access by means of a Web browser to the data over the
Internet. The user can specify trend plots and data tables for any time period
using Vista Data Vision. The software continuously monitors the status of
specified fields in the database and is programmed to send e-mail and text
messages when flow values exceed specified set points. This provides a means to
notify field personnel when runoff from a storm exceeds a particular level
(e.g., 4 inches and 8 inches). We set up the system to notify users via e-mail
or cell phone when water levels have crossed these trigger
points.
The
alarm and notification features are important because supplemental sampling for
oil and grease and for E.
coli
are required at the Halliday site and its accompanying down river sampling
location, as well as at other locations throughout the storm sewer system and
their companion river sampling locations. The alarm also functions as a general
stormwater event monitor, allowing for notification of city staff during storm
events and facilitating sampling of other stormwater outfalls as required by the
permit. Sampling is conducted by city staff at all the required stormwater and
river monitoring stations as long as stormwater flow is
continuous.
|
| Figure 3. Water discharge from the Halliday storm drain to the Portneuf River |
Approach
As
part of the process to determine total annual discharge and develop an estimate
of the pollutant load for each of the pollutants of concern, the city selected
the Halliday Street storm drain for our initial sampling effort. Halliday is the
largest stormwater pipe, has easy access at the end of a city street, and has a
readily accessible electric power supply. The Halliday Street drain is unique in
its large size but otherwise is a representative stormwater outfall. The
Halliday Street drain has a steep slope in the upper reaches and a slope of
approximately 0.001 in the last 800 feet. The total length of the system is over
10,500 feet from the east side of Pocatello to its discharge point into the
Portneuf River in the concrete channel. Twelve feet from the discharge point,
the 84-inch pipe is discharged into a 6-foot by 10-foot grated vault. Flows from
the vault enter a 6-foot section of 84-inch pipe for discharge into the channel
(Figure 3). PUA permit conditions allow for either automatic sampling or grab
samples for stormwater for pollutant load characterization. We elected to
instrument the Halliday Street storm drain so that it would function as a storm
event alarm for flows throughout the entire system.
To
calibrate the Halliday stormwater outfall and allow for calculation of total
flow and total pollutant load, city staff developed a calibration flow protocol
using an offline city well. City Water Department staff estimated that, by using
the three available booster pumps, water could be routed from a 3 million gallon
water-storage tank and pump a total of 6,000 gallons per minute (gpm) through
the storm drain. To enable the city to accommodate the anticipated flows and
pressures during the calibration event, Rain for Rent (Idaho Falls, ID) was
contracted to provide high-density polyethylene (HDPE) pipe and fittings. HDPE
was selected because of its flexibility, a pressure rating of 160 pounds per
square inch (psi), corrosion resistance, zero leakage, extreme durability, and
resistance to weather, chemicals, fatigue, and surges.
The
booster pump was isolated from the rest of the municipal potable water supply by
valving so all the water measured at the booster was directed to the Halliday
stormwater drain. During the morning of the test, it was fairly cold, and there
had been no precipitation during the previous 24 hours, ensuring that the
measurements of flow were not compromised by outside flow influences.
|
| Figure 4. HDPE Pipe assembly at the Halliday test site |
|
| Figure 5. HDPE field assembly and tie down |
The
HDPE was premeasured and fused by Rain for Rent at Idaho Falls for assembly on
the job site (Figure 4). The pipe was heat fused with the Tracstar 500 Fusion
Machine. Each of the six flange adapters was rated to 150 psi and individually
fused to the lengths of pipe. Each fuse took approximately 30 minutes. At 6,000
gpm, the water was experiencing only 2.5 psi of friction loss through the 93.5
feet of 12-inch HDPE pipe. Water velocity was at 19.5 feet per second.
Increasing amounts of water were pumped to the storm drain until the 6,000-gpm
capacity was reached. Measurements of flow at the pump and in the vault were
taken as flow increased and as flows were subsequently decreased. The pipe was
held in place by several cubic yards of road base (Figure
5).
Results
and Discussion
We
used Manning’s equation for flow in open pipe to correlate estimated flow rates
from the booster pumps with the depth of water flowing in the pipe as measured
by the Milltronics equipment. Assuming a Manning’s n
of 0.010 and the known pipe slope of 0.001 the results were as shown in Table
1.
As
is evident from the data, the calculated flow rate is approximately 10% below
the flow rate recorded at the booster pump. Several variables may account for
all or a portion of this discrepancy:
-
The accuracy of the flow meter at the pump station is unknown, leading to
potential errors in measuring actual flows.
- 6-foot by 10-foot vault near the outfall end of the Halliday Street storm
sewer complicated accurate measurement of calibration flows. This change in
conveyance configuration shape and slope may have complicated accurate
measurements of depth of flow. Our depth measurements with the Milltronics also
varied due to wave action in the vault and/or the stability of the sensor.
Graphic presentation of the various data components is provided in Figure 6. We
calculated an average depth to use in the equation.
- We assumed a Manning’s n
of 0.010 for smooth concrete pipe.
Even
with all the variables in play, the comparison of the data showed the flow rate
tracking parallel to the curve predicted by the Manning equation. We intend to
use the Manning equation with an n factor of 0.010 to calculate our flow rates
and total flow during storm events, because we believe that as the depth of flow
in the pipe increases at higher flow rates, the effect of the variables noted
above will diminish. During the calibration test, we noticed that the Miltronics
level transducer, which was suspended from the grating with a cable, was swaying
somewhat with the wind. We subsequently changed the transducer mounting in the
vault to a solid metal arm, which appears to have reduced the variability in
level readings.
The
monitoring installation at the Halliday storm drain will allow us to calculate
total flow during individual storm events and subsequently calculate the total
load for the constituents of concern (sediment, phosphorus, nitrogen, oil and
grease, and bacteria) using an integral of the area under the flow meter curve.
The calibration event undertaken by the city bolsters our confidence in the
accuracy of the flow data used in our stormwater management program.