2. What does a Stormwater Utility do?
• Road
drainage
Maintenance,
repair & construction
of stormwater pipes
• Dams
Prevent downstream flooding
• Water Quality
Regulatory compliance &
reducing pollution
5. Water
Federal EPA
(Agency)
NPDES permit (1990)
Large cities > 100K
Nonpoint Source Pollution
Improve Watershed
Reduce Pollutant Loadings - TSS
(Fecal Coliform, Biota – Fish & Macroinvertebrates)
Federal Clean
Water Act of 1987
And Amendments
State EPD
(Division of DNR)
County and
Municipal
Stormwater
Plans
6. Now (2006 & Forward)
• Proactive Utility
• Technology
7. Decisions, Decisions
• Identify most vulnerable assets
Through critical analysis.
Criteria were identified through brainstorming.
What data supports the analysis?
Why waste time and resources inspecting good features?
8. Who gets involved?
• WIP
• DAMS
• Construction plans
• Maintenance
•Compliance (with WQ)
•Inspections (SRs)
•GIS
9. How Do We Decide?
Then: “Respond to Service Requests” & “Do pond inspections”
Now: NPDES 5 year permit from EPD
“Inspect 10% of the total assets”
“Inspect 20% of the most critical over 5 years & repair if needed”
Maximum Cost Benefit
10. What is the most critical?
• Risk Assessment Model
Likelihood of Failure
Consequence of Failure
11. Likelihood
of failure y-axis
Criteria Weight
Corrosive Soils 20
Age, % of useful remaining life 80
Pipe material 100
Inspection results 100
Inspection results: Inventory based on Good/Fair/Poor Objective Visual Assessment Inventory done over several years beginning in 2000
Criteria Scoring ( 0 = low consequence, 10 = high consequence)* factor of ten
12. Consequence
of failure x-axis
Criteria Weight
Diameter 90
Potential for Safety Risk and/or Property Loss
Proximity to buildings
Proximity to Critical Facilities (hospitals, fire stations, etc.)
Proximity to roads and bridges
80
80
80
Proximity to Impaired (303(d)) Streams 40
Proximity to Environmentally Sensitive Areas
Watershed improvement projects
Drinking water sources
30
30
Criteria Scoring ( 0 = low consequence, 10 = high consequence)* factor of ten
21. Assuming You Have Stormwater Inventory Data…
2000 2002 2004 2006 2008 2010
Whole
subdivisions
Last
Contract
Deliveries
GPS in-house
Digitized from
plats/fieldwork
Ponds
Contracted Out
Inventory/Assessment
Assuming You
Have Stormwater
Inventory Data…
26. Create new field
RUL_score
Select records
Perc_RUL <= 20
Field calculate RUL_score
10 * 80 (weight*factor) = 800
Complete the remaining records
by selected ranges
Finish Loading the RUL_score Data Fields With Final Scores Based on Weight
27. Create new field
Material_S
Select records
“PL” or “CO”
Field calculate Material_s
10 * 100 (weight factor) = 1000
Complete the remaining records
by selected ranges
Finish Loading the Material_s Data Fields with Final Scores Based on Weight
28. Build Consequences
All consequences are related to proximity
Except for Pipe Diameter
Use the “Join data from another layer based on spatial location”. “Minimum”.
30. Graph the results in the *.mxd …
You can select the
most critical
likelihood and
consequence
scores in the graph
and see them in
the view
Graph the resulting matrix in the *.mxd Tools>Graphs>Create>Scatterplot
35. 800 of the most critical pipes
Results
(static
image)
36. Annual Pipe Inspections
Total pipe assets
20%
20%
20%
20%
20% annual inspection
Total pipe assets
10%
10%
10%
10%
10%
10%
10%
10%
10%
10%
Critical pipes thoroughly pole camera or
video inspected annually – PACPS.
Within 5 years, all the most critical will
have been inspected.
Pre-Suggested by EPD Approved with SAMP
The critical pipes happened to be
5% of the total pipe mile assets.
37. EPD Approved!
This is the first known SAMP for SEPARATE (MS4) stormwater system in the USA
38. Critical Pipes With Contiguous Inspection
Critical Pipes
with
Contiguous
Inspection
42. The Cycle of Review
What did inspections reveal?
Should we revise criteria for rating pipes?
Is the SAMP
working?
Do we re-run the
matrix with refreshed scores?
Are there other
consequences
to consider?
43. Contacts
www.gwinnettstormwater.com
678-376-7193
• Steve Hart, Planning Manager
Steve.Hart@gwinnettcounty.com
678-376-6925
• Karen Lougee, GIS III
Karen.Lougee@gwinnettcounty.com
678-376-6719
• John Dean, GIS III
John.Dean@gwinnettcounty.com
678-376-6761
Editor's Notes
Objective is to show how to do asset management with GIS of a pipe conduit system using a matrix with X and Y values.
Road Drainage: 1300 miles of Stormwater pipe ranging in size from 12 inches to 108 inches. 140 miles of impacted streams. Dams: 15 cat. 1 Water Quality: NPDES permit, (5 years) MS4. National Pollutant Discharge Elimination System. Municipal separate storm sewer system. A conveyance or system of conveyances (including roads with drainage systems, municipal streets, catch basins, curbs, gutters, ditches, manmade channels, or storm drains) owned by a state, city, town or other public body, that is designed or used for collecting or conveying storm water, which is not a combined sewer, and which is not part of a publicly owned treatment works. Commonly referred to as an &quot;MS4“. Dam Cat 1: &quot;any dam located as to endanger populated areas downstream by its sudden failure.&quot; Could cause substantial loss of life, excessive economic loss, or substantial public inconvenience by its sudden failure. WQ:EPD.
Reactive: Responds to service requests incoming: Repairs, Crisis management. BMP Inspections of ponds. Funding: What do you do when the money runs out? 30 year backlog of maintenance issues
Responses to incoming calls, and inspect BMPs.
From 118 outfalls to 2,500 outfalls by new definitions. Consequences: $25,000 fine per day per outfall. Clean Water Act. Point sources permitted. Nonpoint is stormwater. Federal Clean Water Act of 1987 required EPA. NPDES in 1990. National Pollutant Discharge Elimination System. We (Gwinnett County) are in phase 1 due to size (pop &gt; 100,000 people) Locally – nonpoint source pollution. Polluted stormwater runoff is commonly transported through Municipal Separate Storm Sewer Systems (MS4s), from which it is often discharged untreated into local waterbodies. To prevent harmful pollutants from being washed or dumped into an MS4, operators must obtain a NPDES permit and develop a stormwater management program. Phase I, issued in 1990, requires medium and large cities or certain counties with populations of 100,000 or more to obtain NPDES permit coverage for their stormwater discharges.
Self-funded (impervious billing) known revenue stream to plan budget ahead: prevention, extend life of assets. Technology – CIPL, GIS, PACPS. Inspect 10% of everything. Inspect 5% of the top 20% critical each year. Extend useful life of pipe assets with CIPL and PACPS (Pipeline Assessment Certification Program).
Likelihood vs consequence. Why waste time and resources inspecting good features just to meet a suggested “inspection” goal. 2 hours each Friday x 6 meetings. Identify what streams to clean up 303d stream list. Pollutants. Does stream meet intended use? Fishing, recreation, drinking water.
Watershed Improvement Projects – Pete Wright/constructed wetlands and stream restorations. Dams – Cat 1 – Jonathan Semerjian. Construction & Maintenance – John Corn. Compliance /Illicit Discharge & TMDL stream monitoring. Inspections of BMPS – Brian Speers GIS – data collection and editing
EPD 5 year permit. We have 4080 critical pipes in the highest 3x2 grid. 20% of most critical is 816 pipes per year to be inspected.
Likelihood – quantitative. Consequence – qualitative. And the most benefit. Bang for the buck.
Inspection results Inventory based on Good/Fair/Poor Objective Visual Assessment Inventory done over several years beginning in 2000. More quantitative data. Weight is the RAW SCORE * Factor of TEN.
More qualitative data. Weight is the RAW SCORE * Factor of TEN.
Likelihood – iterative process
Consequence – iterative process
Final Likelihood weights are still a work-in-progress
Final Consequence weights are still a work-in-progress
Gather raw data. Clean it. Build ranges.
Effect on pipes: steel rusts but pvc & concrete does not. Weight factor 2 * ten.
Use your data dictionary!
Manufacturer specifications.
STW pipes inventory collected from plats – digitized field work – from benchmarks – 2000. 2001 – started doing GPS inhouse. Ponds first. Then in 2002 whole subdivisions. Some contracted out. Last delivered end 2006.
City of Lilburn joined our STW Utility , Y1 = Lilburn pipes. Y = County Maintained Pipes.
All Likelihoods are quantifiable.
Then = construction date. Now = current year. Diff = now minus then (which is age of pipe). Material life = manufacturer specs.
Remaining life is manufacturer’s “expected life” minus “age in years” of pipe.
Finish loading Perc_RUL field using the weight factors for each range
PL = Plated Corrugated Metal, CO = Corrugated Metal are score value of ten.
Material score value of 10, 8, 6, 4, 2 depending on type. Multiply each type in a selected range by weight of ten.
BIG PIPES = BIG CONSEQUENCES. All proximity builds are buffers from spatial locations. Score based on established weights. Hospitals, Police Precincts, Fire Stations. Merge them. Build concentric buffers of 5 rings 100 feet apart. Select pipes that intersect the buffer & export them. Use the “Join data from another layer based on spatial location”. That will assign distance attribute from the buffer to the pipe. check “minimum”. Reality? Pipes we maintain are not that close.
Add the scores from the fields you built for the X values and add the scores from the fields you built for the Y values. Use field “likelyscor” for y-axis and field “conscore” for x-axis
Tally scores for “Likelihood” and “Consequence” .
RAW Quantitative evaluation (not using SAMP)
Qualitative evaluation
If we had been a self-funded utility 5 or more years ago perhaps we could have prevented “some” of this 500-Year Flood issues. We did not have the budget to do it. We did not have GIS technology trained people or other resources such as mapping of the infrastructure. We could not anticipate under-sized pipes! Qualitative variables have a greater bearing on the outcome due to their influence on other parameters. Quantitative variables are purely mathematical. A conclusion drawn from a mix of highly qualitative variables is more likely to be predictive (correct) even if their numeric value is not as widespread.
Showing how to build the matrix using a graph tool in GIS
This sample “how to build a graph” does not indicate the 20% grid ranges that our model uses. Because we did not want to redo this demo sample!
We will Inspect 10% of everything (82K). Inspect 5% (that is 800 pipes) of (a selected range in) the top 20% critical each year. (5 years to inspect those selected in top 20% critical).
Note this difference here of number of pipes vs miles of pipes. Two different things. OUR STORMWATER MANAGEMENT PLAN TO STATE FOR THE NPDES PERMIT IS REPORTING IN NUMBER OF MILES OF PIPES.
This is the first known SAMP for SEPARATE (MS4) stormwater system in the USA. (Seattle is a combined system). We will Inspect 10% of everything. (80K) Inspect 5% (that is 816 pipes) of the top 20% critical each year. (4K pipes). Discuss why we don’t inspect/repair 100% of all the most critical pipes in the first years.
2 critical pipes with the contiguous pipes and structures to be inspected. Budget plus “maximizing cost benefit” = why. If these pipes score “4” or higher go on to the next contiguous pipe or structure until reaching a “3”. Using NASSCO’s PACPS (pipeline assessment and certification program). This builds out to the 10% of pipes we report in our stormwater management plan. INSTEAD OF DOING A CURSORY INSPECTION OF ALL OUR ASSETS IN 5 YEARS, WE WILL DO A THOROUGH INSPECTION USING POLE CAMERAS AND PACPS ON A SMALLER, SELECTED CRITICAL NUMBER OF MILES OF PIPES.
Cannot anticipate undersized pipes.
500 year flood in September 2009. We could not have prevented these issues due to lack of resources.
Review parameters are not established. It is a work-in-progress. Plain End pipes issues. PLAIN END PIPES AS AN INLET FEATURE ARE VIRTUAL!
