riverspill and the incident command tool for drinking water … · 2020. 4. 3. · riverspill and...

18
RiverSpill and the Incident Command Tool for Drinking Water Protection A case study of the WV chemical spill William B. Samuels, Ph.D Leidos - Center for Water Science and Engineering USGS Hydrology Seminar April 9, 2015 OHIO hio Ri er

Upload: others

Post on 28-Feb-2021

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

RiverSpill and the Incident Command Tool for Drinking Water Protection

A case study of the WV chemical spill

William B. Samuels, Ph.D Leidos - Center for Water Science and Engineering

USGS Hydrology Seminar

April 9, 2015

OHIO

hio Ri er

Page 2: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Drinking Water Protection

• After a release of hazardous materials into a river or stream environment, drinking water protection and contamination risk mitigation require that information on the fate of waterborne contaminants be made available quickly to decision makers.

• On January 9, 2014, an estimated 10,000 gallons of 4-methylcyclohexane methanol (MCHM), which is used in coal processing, leaked from a ruptured container into the Elk River.

• The spill, just one mile upstream from a water-treatment plant, forced officials to ban residents and businesses in nine West Virginia counties from drinking the water.

• Downstream water utilities were also concerned about potential contamination.

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 2

Page 3: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

0 2& MILES

Ka11m,:haR.

◄ lJ

Rivcr,~ -.Jll<Ytv

C1hia Ri""

W EST VIRGINIA

Area of tap wat er

,-- use ban

•Beckley

71

VI GIN A ~-~

Chemical spill threatens thousands A chemical, 4-methylcyclohexane methanol, a foaming agent , used in the coal-preparation process, leaked from a tank at Freedom Industries into the Elk River upstream from Charleston on Thursday. Read related article.

.."'.,... .. 3000

I I I ' FEET

... 1 ~,--~ • Total population: 294,350; of that 65,237 are chi ldren • 123,451 households • Median household income is $47,191 (W.Va. average is $40,400) • Number of families: 78,856; of those 8,445 in poverty; poverty rate 10. 7 percent (W. Va. average is 12.8 percent)

Source Google Earth and the US Census Bureau Gene Thorpe/The Washington Posl Published on January 10 2014. 7 45 pm

Population affected:

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 3

Page 4: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Chemical Storage Tanks

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 4

Page 5: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

RiverSpill - ICWater

Input Databases User Inputs

Location

Source Term NHDPlus

USGS Real-Time

Gages ArcGIS . .

Assets RiverSpill

Contaminant Database

Outputs

Downstream Trace

Breakthrough Curve

Upstream Trace

• Simulate fate and transport of the contaminants from point and non-point sources,

• Use minimum input data and user interface interaction for rapid emergency response modeling of toxic spills

• Perform hydraulic transport for the entire US river system

• Select water quality state variables that can represent the major processes that control water quality impacts and parameters available from national databases.

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 5

Page 6: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

RiverSpill - IC ater

LOOP I P.1US:£ J PlAY / QI/TT J

' .<n_, Soeod (Fr"".., o,, Seconc!)

,- F.,t'10!ooJ r. Hormol (Sii>o) ,- Slow /I fool

0 62Shcu, 2.7042iog,1

0 (.I\

0 vl 0 ~

0-T

~

~--

O S"O 10 0 09

P e n.,sy 1a n1 a

u5u20005

Where is the contaminant going?

• Is there a drinking water intake in its path?

• When will it reach drinking water?

_____...,.. .I

j • Is its level high enough to be a human threat?

©2015 LEIDOS. ALL RIGHTS RESERVED. ~ leidos 6

Page 7: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

RiverSpill - ICWater Modeling Principles

• Source Term Instantaneous or continuous release

• Mixing Instantaneous and complete mixing in the water column

• Velocity NHDPlus mean flow and velocity updated with USGS real-time gages

• Dispersion 1-D longitudinal dispersion

• Decay First order exponential decay

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 7

Page 8: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

41200.000

2340

-

West Vir ·n Kana,vh

► - ►-

,..

.t ►-

cfs Velocity.

Velocity Factor.

Selected Gage·

►-

3.149

1 260

Realine Vefociy NHD Row

N/ A 17606.42987

►-

►-► - ◊

J> , ►

ES 0

NHOVeJociy

2 50441

fps

►-

Temperature 20

Row Factor

N/ A

VeJ Factor

N/A

itream velocity, feet per second Measured Discharge, cubic feet per secondMost recent instantaneous valu : 3 .07 0 -13- 2014 11:00 EST

velocity Most recent instantaneous value: 41,200 01-13-2014 11:00 EST USGS 03198000 KANAMHA RIVER AT CHARLESTON, M

3 _5 ,---------------------------, USGS 83198888 KANAMHA RIVER AT CHARLESTON, MV

.,,,6 3 .0 .. -c, 48888"

" C I" 0..c.. 2.5 Cl) ..Q.

:>ws Help " :l 2. 0 ~ 38888 ... "' I:L I~ CDJ ~ Ii! l!J I~ ~ l ICWater• I +>

0..

• -• Cl) I., "' 1.5 ...Cl)....

0 28888" .... .... ~ 1 . 0 " .&J ::,s: ".."'

_;; 0.5 t' 18888., ~"' c..

o.o '---------------------------' Jan Jan Jan Jan Jan Jan Jan Jan ." " ... 006 07 08 09 10 11 12 13 =

2014 2014 201 4 2014 201 4 2014 2014 2014

-- Provisional Data Subject to Revision---­Jan Jan Jan Jan Jan Jan Jan Jan

@ U HALMA 1: Hazaraous \IYaste :>1tes 86 87 88 09 10 11 12 13 2814 2814 2014 2014 2014 2814 2014 20148 0 Dischargers: Municipal & lndustr

---- Provis ional Data Subject to Revision ----fJ Industrial Facilities ►

-., -1J Sewage Treatment Plants

(!J 0 Public_Water_Supplies: Intakes 8 0 Scenario Layer

8 0 Point Sources

•B 0 Polygon Sources

D B 0 ICWater Base Map

ffi O Transportation Lines

B 0 NHD @ 0 NHDPlusOS

(±l 0 Transportation Polygons @ 0 North America

@ EJ World

Edit Flow Factor

l.lsetSpeotied

Flow:

Flow Factor:

USGSReabne

Gage Name

KANAWHAR

KANAY./HA RI►

Page 9: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

uses 03255000 OHIO RNER AT CIHCIHHATl, OH ..

OHIO

USGS 03206000 OHIO RIVER AT HUNTINGTON, WV .... r---------------------, 3J.8

..... Jin Jan ...,. -"-

.. u u u •• 281 ◄ 281 ◄ 291 ◄ 281◄ 20!◄ - ,r,OVi.sion.aJ, a.te .S..J.ct to llevist., -

WEST VIRGINIA

.,,....__ . -

Flow Regime

.. ! ~

~ • 34

'£-j ~ 32

.. 30

-.. .,_ ...... .,_ ...... ...... .Jan ~ .Jan.... u u u u 1◄ 1, a u291◄ 28.1◄ 28.1◄ 291.◄ 211.l◄ 2111.◄ Ml◄ 2at◄ - ""9vldon.J kt.. Sub~t to lt~b:Jon - -

.... ....<------------------::---:---~•:• u:ee ee:oe 12:ee oe:ea 12:ee ee:ee t t.:•

J_,99 -.99 .,_.u ....,18 -.u .J.wtu .i.nu .Jarttt28.l◄ -◄ at◄ 28J◄ 281◄ 2eJ◄ 291◄281A _,._ ,.,..OYhionol O.t• Sub.Ject to t.YJ.aion -

t.. I

9

©2015 LEIDOS. ALL RIGHTS RESERVED. ~ leidos

Page 10: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

I

Downstream trace

ICWater 3

File Edit View Bookmarks Insert Selection Geoprocessing Customize Windows Help

i ~ ~ ,tl\ ~" ) D ~ (ill cgi I -llw "") ~ ,( I r <:!;, T I 1:83.020 .., :,L I GJ Cj ~ Iii l!'.J I ~ .: i ICWater T I Scenario T Downstream T Upstream T Reports T I Rainfall T I Flow Conditions III I • IL I ?' I L.Ji'

q. X r--------==:----r-.:--==------:::;::::..-----,7,.....c.- -----......,::::::::7----------,c_-,_,7--- --r----:~-------------=::=-- ...Table Of Contents

~ g ~ .@. I ~ Putnam 8 ~ ICWater Analysis

l±J ~ Gages: Realtime List of Water Systems in SDWIS 8 ~ Trace End

0 I To obtain additional information about drinking water please call EPA's Safe Drinking Water hotline at 1-800-426-4791.

8 ~ Downstream Trace •

AboveLOC Community Water Systems: Water Systems that serve the same people year-round (e.g. in homes or businesses).

Transition Zone Below LOC

- - Above LOC, Tidal Region Water System Name County(s) Served Population Served Primary Water Source Type System Status Date Closed Water System ID

WVAWC-KANAWHA VALLEY DIST KANAWHA 217959 Surface_water Ac1i11e WV3302016 •- -. Transition Zone, Tidal Region - - Below LOC, Tidal Region :J

8 ~ Scenario Layer

t-,_<..!!====== _.) ~ €1 I ~ uII (

8 ~ Point Sources

•8 ~ Polygon Sources

□8 ~ Critical Customers: Schools ►-

8 ~ Critical Customers: Hospitals

r::J @ D HAZMAT: Risk Management Plar

l±l D HAZMAT: Toxic Chemical Releas

l±l D HAZMAT: Hazardous Waste Sites

8 ~ Dischargers: Municipal & Industr

I) Industrial Facilities

fJ Sewage Treatment Plants

(tJ ~ Public_Water_Supplies: Intakes (tJ ~ ICWater Base Map

IJ irginia

SsintAl!>ans

- ►

-►Duaba

► r::J - ►

l

-

l)IJ !

►-

►-►-

,..-

r::J

"'=====-'---~

WVAWC- KANAWHA VALLEY DIST

585 GLADY FORK ROAD

WESTON, WV 26452

304-269-2006

►-

►-

...

81°47'5797"W 38°23'11.42"N Primary Water Source Type Population Served

Surface_water 217959

Page 11: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

- -

l

24 hour downstream trace File Edit View Bookmarks Insert

q. XTable Of Contents

El ~ ICWater Analysis El 0 Trace End

0 El 0 Downstream Trace

Above LOC Transition Zone Below LOC

- - Above LOC, Tidal Region • • Transition Zone, Tidal Region - - Below LOC, Tidal Region

El 0 Scenario Layer El 0 Point Sources

•El 0 Polygon Sources

□l±J 0 Critical Customers: Schools 1B 0 Critical Customers: Hospitals l±J 0 HAZMAT: Risk Management Plar l±J 0 HAZMAT: Toxic Chemical Releas 1B O HAZMAT: Hazardous Waste Sites EJ 0 Dischargers: Municipal & lndustr

I) Industrial Facilit ies

I) Sewage Treatment Plants

l±J 0 Public_Water_Supplie.s: Intakes El 0 ICWater Base Map

1B O Transportation Lines

El 0 NHD ffi 0 NHDPlusOS

ffi 0 Transportation Polygons ID 0 North America ffi @ World

Selection Geoprocessing Customize Windows Help

r 1 ¢> . 1 1:166.936 .., I ~ GI wJ l:8) liJ [!:1 I ~ ( ICWater • Scenario • Downstream • Upstream • 1 Reports • I Rainfall • I Flow Conditions I Iii) I • [QI i I 4-~ ...

<Jackson

-, -

,.

◊ ►~i

\ ►

► -'

~ -7►

.t:~ ►

l...◄-!:I====="=' ====~I___►_J ~ rll I ::J II < u ►

s1°46'54.ss•w 38°34'57.0S"N

Page 12: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

1.. -

Cincinnati Observed Data ICWater Simulated Results

Time 148 hours from the spill 150 Hours

Maximum 20 ppb 14 ppb Concentration

Charleston Observed Data ICWater Simulated Results

Time 21 hours from the spill 27 Hours

M aximum

Concentrat ion

1.87 -3.17 ppm * 5.5 ppm

*Two samples were tested at two different lab locations - WV Division of •

Homeland Security and Emergency Management27

Huntington Observed Data ICWater Simulated Results

Time 90 hours f rom the spill ,85 Hours

Maximum

Concentration

73 ppb 74 ppb

.J

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED.

1

12

Page 13: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Cincinnati 0 .025 ------------------------------

0.02 -+--------------------------------------

-E Q. Q. "; 0 .015 .Q.. L!..8 0 .01 C

8 0 .005 --------

0 +------

♦ •• •♦ ♦

110 120 130 140 150 160 170 180 190

Time (hours)

• Simulated Cone (ppm) ♦ Observed Cone (ppm)

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 13

Page 14: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Challenges • A major challenge in modeling this spill was the characterization of the source

term. Uncertainty of input parameters: - Spill duration

- Release pattern

- Volume released

• Toxicity levels of MCHM

• Accounting for how much mass was lost through downstream intakes

• Accuracy of MCHM measurements for model updates and comparisons

• In this particular spill, an added challenge was that the pollutant's delivery was attenuated by the soil that much of it had to seep through before reaching the

•river

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 14

Page 15: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Success Factors

• The NHDPlus national river network coupled with real-time stream flow data and river forecast data -this allowed for the ICWater model to simulate the leading edge, peak concentration, and trailing edge of the spill from its origin on the Elk River to intakes hundreds of miles downstream.

• Model runs were updated based on MCHM measurements at downstream locations on the Ohio River - this provided accurate forecasts to nearby water intakes.

• GCWW, a large water utility on the Ohio River, used ICWater along with river grab samples, to determine when to close its intake to allow the spill to pass by.

• Data for Cincinnati showed good agreement (within several hours) between the observed peak time of arrival and the model's estimated peak time. The leading- and trailing-edge predictions were also close to the observations.

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 15

Page 16: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Lessons Learned

• Initial spill reports are often inaccurate - constantly update forecasts based on changes in source term and environmental conditions

• Use both real-time gage data and stream flow forecasts

• Base decisions on both model forecasts and field observations -use measurements to update model inputs at intermediate downstream locations

• Teamwork - collaborate with local authorities, water utilities and data providers

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 16

Page 17: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Concl usions

• For an event of this magnitude, this modeling response showed the utility of a national level toxic spill model (as opposed to regional river basin models) to simulate the entire pathway of the spill in a timely manner.

• The integration of a hydrologically connected - national level stream network, real-time steam flow, river forecast data and sampling observations enabled decision makers at a major water utility to take appropriate action to protect its water supply.

~ leidos ©2015 LEIDOS. ALL RIGHTS RESERVED. 17

Page 18: RiverSpill and the Incident Command Tool for Drinking Water … · 2020. 4. 3. · RiverSpill and the Incident Command Tool for . Drinking Water Protection A case study of the WV

Further Reading

Bahadur, R. and Samuels, W. (2014). ''Modeling the Fate and Transport of a Chemical Spill in the Elk River, West Virginia." J. Environ. Eng., 10.1061 /(ASCE)EE.1943-7870.0000930

Samuels, W.B., Bahadur, R., Ziemniak, C. and Amstutz, D.E. (2014). ''Development and Application of the Incident Command Tool for Drinking Water Protection'', Water and Environment Journal, http://onlinelibrary.wiley.com/doi/10.1111/wej.12097 /abstract

~ leidos ©2013 LEIDOS. ALL RIGHTS RESERVED. 18