Battelle Chlorinated Conference, 2018
®
Per Loll R&D Manager, Ph.D.
Sewers as a Preferential VI
Pathway – Dynamic
Measurements and
Quantitative Risk
Assessments
Poul Larsen and Claus Larsen, DMRHanne Nielsen, Kristian Raun, Kim Thygesen and Klaus Mortensen, Region of Southern Denmark
®
2Per LollBattelle Chlorinated Conference, 2018
Preferential pathways
• A preferential pathway serves as a high permeability and capacity pathway of VOC vapors from a source area to and into a building.
• We distinguish between external and internal preferential pathways.
– External: Sewers, land drains, utility tunnels, …
– Internal: Cavity walls, elevator shafts, stairwells, attic spaces, …
®
3Per LollBattelle Chlorinated Conference, 2018
Some questions about sewers at any given site
• What can be expected from the sewers at a given site?
– Size of differential pressure and concentrations?
– Temporal variation in differential pressure and concentrations?
– Size of sewer air exchange (design of tracer gas studies)?
– What is the effect of mechanical ventilation (fan)?
– What is the effect of a hole in the sewer?
– How to deal with 2-4 OOM temporal variability?
– How can we make sure that we collect samples at the “right” time for a robust risk assessment?
• Standard Danish practice in relation to risk assess-ment is 10 L samples of sewer air on Dräger B carbon tubes for lab. analysis (could be Summa).
• What to expect? Tools and thoughts on application.
®
4Per LollBattelle Chlorinated Conference, 2018
• We looked at 2
The test site in the Region of Southern Denmark
• A three-level building from 1900.
• Apartments on the 2nd and 3rd floor.
• Various activities on 1st floor through time
– Plumber
– Electronics Repairs
– Foot Clinic
• Site investigations and remediation attempts 2000-15.
• Building was condemned and was purchased for tests.
– Dry Cleaner (PCE)
1st floor 2nd floor
3rd floor
• 6 toilets in all.
®
5Per LollBattelle Chlorinated Conference, 2018
• Logging of proxy and secondary parameters + automated sampling for chemical analysis (via valves and pumps):
We created an automatic sampling unit (prototype)
PID-sensor– 2 PID-sensors (10,6 eV; 1 ppb-40 ppm & 0,1-6000 ppm)
– 2 differential pressure (dP) sensors
– 1 barometric pressure sensor
– 1 temperature sensor
PID-sensor (room)PID-sensor (sewer)
dP sensor (sewer)
dP sensor (floor/outer wall)
2 x carbon tubes for sample collection
LED Display
– Option of collecting 2 samples on carbon tubes
– Display, on-line data logging & remote controllable
®
6Per LollBattelle Chlorinated Conference, 2018
Sampling toilet (for equipment protection)
• We constructed a sampling toilet which allows for regular use while logging (shown for the 1st floor toilet):
– Constructed using standard plumbing components.
• The actual toilet was leak tested (no leaks).
• So was the sampling toilet (no leaks).
• 5% H2 / 95% N2
®
7Per LollBattelle Chlorinated Conference, 2018
Differential pressure – 1st floor toilet
• 1 month of logging 4 times per minute.
Door closed, toilet not in use
Toilet
FloorToilet vs. floor
Average 1,5 Pa
Min -5,3 Pa
Maks 7,9 Pa
Percent > 0 Pa 98 %
Percent >+1 Pa 82 %
Percent >+2 Pa 11 %
Average 0,31 Pa
Min -2,3 Pa
Maks 3,1 Pa
Percent > 0 Pa 91 %
Percent >+0,5 Pa 21 %
Percent >+1 Pa 0,12 %
®
8Per LollBattelle Chlorinated Conference, 2018
Average 0,08 Pa
Min -19 Pa
Maks 16 Pa
Percent > 0 Pa 65 %
Percent >+2 Pa 0,9 %
Percent >+5 Pa 0,04 %
Average 0,77 Pa
Min -19 Pa
Maks 17 Pa
Percent > 0 Pa 85 %
Percent >+2 Pa 2,4 %
Percent >+5 Pa 0,05 %
Differential pressure – 2nd floor toilet
• 1 month of logging 4 times per minute.
Door closed, toilet not in use
Toilet
Outer wallToilet vs. outer wall
®
9Per LollBattelle Chlorinated Conference, 2018
Differential pressure – 2nd floor toilet
• Significance of absolute barometric pressure?
– High (>1020 mbar)
– Low (<1005 mbar)
Baro vs. outer wall
Baro vs. toilet
Toilet
Outer wall
Barometric pressure
®
10Per LollBattelle Chlorinated Conference, 2018
R² = 0,0424
-2
-1
0
1
2
-2 -1 0 1 2
dBaro/dt vs. WC (timemidling)
R² = 0,0424
R² = 3E-05
R² = 0,0675
-2
-1
0
1
2
-2 -1 0 1 2
dBaro/dt vs. WC (timemidling)
R² = 0,008
-2
-1
0
1
2
-2 -1 0 1 2
dBaro/dt vs. Ydermur (timemidling)
R² = 0,008
R² = 0,0023
R² = 0,0123
-2
-1
0
1
2
-2 -1 0 1 2
dBaro/dt vs. Ydermur (timemidling)
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1,0
1,5
2,0
01-11 03-11 05-11 07-11 09-11 11-11 13-11 15-11 17-11 19-11 21-11 23-11 25-11 27-11 29-11 01-12
Æn
dri
ng
i bar
om
ete
rtry
k (m
bar
/t)
Trykstigning
Trykfald
Stabilt tryk
Rising
Falling
Stable
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1,0
1,5
2,0
01-11 03-11 05-11 07-11 09-11 11-11 13-11 15-11 17-11 19-11 21-11 23-11 25-11 27-11 29-11 01-12
Æn
dri
ng
i bar
om
ete
rtry
k (m
bar
/t)
Trykstigning (29%)
Trykfald (27%)
Stabilt tryk (44%)
Differential pressure – 2nd floor toilet
• Significance of barometric pressure changes?
– Rising
– Falling
– Stable
• Analysis of hourly avg. data
>0,5 mbar/hr
<-0,5 mbar/hr
[-0,5;+0,5] mbar/hr
dBaro/dt vs. toilet
dBaro/dt vs. outer wall
Rising (29%)
Falling (27%)
Stable (44%)
Toilet Outer wall
Barometric pressure
®
11Per LollBattelle Chlorinated Conference, 2018
-7
-5
-3
-1
1
3
5
13:30 14:00 14:30 15:00
dP
(P
a)
Door open
Fan on (door closed)
Effect of mechanical ventilation (1st floor toilet)
Ventilation fan on(closed door):
5-6 Pa negative dP in room
6-8 Pa positive dP across sewer
4-5 Pa positive dP across floor-3
-1
1
3
5
7
9
13:30 14:00 14:30 15:00
dP
(P
a) Baseline (dør åben) In-line ventilator / Strip på dør
dP faldstammedP over gulv
Baseline in agreement with long term data:
1-2 Pa positive dP across sewer
0-1 Pa positive dP across floor
• In-line ventilation fan (~51 m3/hr; ACH= 8,6 hr-1)
– Differential pressure (room) door closed
Sewer
Floor
Fan on (door closed)Baseline (door open)
Ventilation fan adds about 4-6 Pa of extra pressure gradient
®
12Per LollBattelle Chlorinated Conference, 2018
0
100
200
300
400
500
600
700
800
900
10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00
PID
(p
pb
)
Car
bo
n tu
be
sam
pli
ng
(10L
)
+hole
Op
en
to in
stal
l PID
se
nso
r
Car
bo
n tu
be
sam
pli
ng
(10L
)
Car
bo
n tu
be
sam
pli
ng
(10L
)
-3
-1
1
3
5
7
9
14:00 15:00 16:00 17:00 18:00 19:00
dP
(P
a)
Sewer
Floor
Door open
Fan on (door closed)
-15
-10
-5
0
5
10
14:00 15:00 16:00 17:00 18:00 19:00
dP
(P
a)
+5 mm hole
Fan on
(door closed)
Door open for
+2 mm hole
Effect of a hole in the sewer (1st floor toilet)
Ventilation fan +2/5 mm hole leads to < 5,5 L/min flow out of the sewer (<1% of the combined flow/ACH).
10 L sampling leads to a temporary drop in PID (700->300 ppb).
Addition of a hole leads to a ‘permanent’ drop in PID (->200 ppb).
• In-line ventilation fan (~51 m3/hr; ACH= 8,6 hr-1) + drilled hole
®
13Per LollBattelle Chlorinated Conference, 2018
PCE 780 840 780
TCE 39 43 39
TVOC 2300 2000 1100
(µg/m3)
PID/concentration in sewer (1st floor toilet)
PCE 780 840 780
TCE 39 43 39
(µg/m3)
-Hole +Hole
70-80% of TVOCs are C6-C10
0
100
200
300
400
500
600
700
800
900
10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00
PID
(p
pb
)
Car
bo
n tu
be
sam
pli
ng
(10L
)
+hole
Op
en
to in
stal
l PID
se
nso
r
Car
bo
n tu
be
sam
pli
ng
(10L
)
Car
bo
n tu
be
sam
pli
ng
(10L
)
• Automated sampling from sewer
– 2 x without a hole (@ PID = 720) and 1 x with a drilled hole
– In-line ventilation fan on (ACH = 8,6 hr-1)
TVOCs have never been investigated at the site before
®
14Per LollBattelle Chlorinated Conference, 2018
-10
-8
-6
-4
-2
0
2
4
6
8
10
0
1
2
3
4
5
6
7
8
9
10
16-11 18-11 20-11 22-11 24-11 26-11 28-11 30-11
Dif
fere
nst
ryk
(P
a)
PID
(pp
m)
Faldstamme
dP faldstamme
Alm. prøvetagning
Alm. prøvetagning
PID/concentration in the sewer (long term)
PCE 75 700 630
TCE 1,2 4,4 35
(µg/m3)
Sewer samples
(10L on carbon tubes)
PCE 75 700 630
TCE 1,2 4,4 35
TVOC 820 6000 < 1000
(µg/m3)
>75% of TVOCs are C6-C10
Red = maximum value
• Automated sampling (1st floor toilet, the least dynamic):
– PID ≥ 5 ppm and dP ≥ 1 Pa (pressure gradient toward room).
Manual
sample
Manual
sample
PID sewer
dP sewer
Automated
sample
®
15Per LollBattelle Chlorinated Conference, 2018
PID/concentration in the sewer (long term)
-10
-8
-6
-4
-2
0
2
4
6
8
10
985
990
995
1000
1005
1010
1015
1020
1025
16-11 18-11 20-11 22-11 24-11 26-11 28-11 30-11
Dif
fere
nst
ryk
(Pa
)
Atm
osf
ære
tryk
(mb
ar)
Atm.tryk
dP faldstamme
Intelligentprøvetagning
Alm. prøvetagning
PCE 410 480 590
TCE 16 15 26
TVOC 1100 3500 1200
(µg/m3)
Sewer (10L on carbon tubes)
Red = maximum value
• Automated sampling (2nd floor toilet, PID yielded only noise):
– Pressure drop > 1 mbar/hr and dP ≥ 1 Pa (pressure gradient toward room).
Automated
sampleManual
sample
Barometric pressure
dP sewer
®
16Per LollBattelle Chlorinated Conference, 2018
Air exchange in the sewer (method & results)?
0
50
100
150
200
250
300
350
400
450
500
16:00 16:30 17:00 17:30 18:00
PID
(pp
m)
WC 1.sal
+100 ppm
+1000 ppm
100 ppm test
y = -4,94x + 3,52
1000 ppm testy = -5,75x + 5,73
0,0
1,0
2,0
3,0
4,0
5,0
6,0
7,0
0,00 0,05 0,10 0,15 0,20 0,25 0,30 0,35
Ln(P
ID)
Tid (timer)
Natural air exchange: Ls= 5-6 hr-1
A previous study [Loll et al. 2015]: Ls= 160 hr-1
Take home message: Hard to guess; but it’s easy to measure.
• Addition of isobutylene (100 and 1000 ppm) and PID decay.
2nd floor toilet
®
17Per LollBattelle Chlorinated Conference, 2018
Summary
• We have developed and applied an automated sampling unit that allows us to analyze temporal variability and collect samples at desired times (various sensors can be attached).
100 ppm testy = -4,94x + 3,52
1000 ppm testy = -5,75x + 5,73
0,0
1,0
2,0
3,0
4,0
5,0
6,0
7,0
0,000,05
0,100,15
0,200,25
0,300,35
Ln(P
ID)
Tid (timer)
• We also tested:– A sampling toilet that allows for long-term monitoring in normal use.
– A metod for estimating air exchange in sewers (use in tracer studies).
• Such tools allow us to account for high temporal variability in our risk assessments.
– And we found a low level ppb sensor (10,6 eV & 0,5 ppb-4 ppm):
®
18Per LollBattelle Chlorinated Conference, 2018
How to apply in quantitative risk assessment?
• Until we know sewer dynamics better we might need to measure and analyze every time – establish normal and critical behavior.
• Collect samples for lab analysis at critical times + documentation.
– What’s the concentration and how (much) does it vary?
• Determine sewer ACH for design of tracer studies (PFT tracers)
– Then we can determine a quantitative sewer contribution to indoor air.
– I have previously tried designing by guessing at ACH – unsuccessfully.
PID sewer
dP sewer