Typical Equilibrator Design

1

• Typical Equilibrator Design
• Circulating 1.1 shower head or gas/liquid mixing
  device
• Conventional shower head, KO, KEEC
• Combined bubblingmixing device, NOAA, UK, GE, NZ
• 1.2 CO2 permeable membrane
• MBARI Underway, NRIFS
• Flow through 2.1 bubling
• MBARI Mooring
• 2.2 modified bubbling
• NIES Tandem (Patented)
• 2.3 long membrane
• Chemistry sensor pH sensitive dye
• CARIOCA, SAMI
• Red are the 2003 inter-comparison participated.

2
NIES systems (Tandem, bubbling/mixer two stage
equilibrator and automated NDIR) were very stable
and two similar systems gave very small
difference (approximately 0.5 matm difference).
Reference system for further analysis
Hazaki 2003 inter-comparison might not be
appropriate set up for chemistry sensors. High
organic matter content of used coastal
seawater. Discussion will be made in this
workshop.
3
What we learned about the possible error 1 Large
equilibrator gave small temperature difference.
Small equilibrator needs accurate temperature
correction. Underwater pump with 300L/min flow
rate gave temperature increase about 0.05 degree.
It infers on board pump with small flow rate
will give larger temperature increase (electric
heating per water flow rate). Positive bias and
common bias in all on board installation
Solution Thermal insulation especially for small
equilibrators Precise temperature correction, but
it is achievable. Sea chest thermometer have to
be prior to the water pump.
4
Our Pyxis installation, temperature sensors and
water pump
Sea
Water flow
5
Development of New pCO2 system (Flow through- two
stage equilibrator, Tandem)
Tandem equilibrator (two stage equilibrator) 1st
stage bubbling 2nd stage static mixier
We get confidence about the discrepancy of
bubbling equilibrator by the 2nd pool
test Comparison with other systems Theoretical
consideration by efficiency test Surface tension
effect gives 0.8 low bias.
Patent for 3 items 2 stage equilibrator
design flow through pCO2 measurement feed back
air supply to equilibrator
6
Merit of Tandem equilibratorIt was very stable
and reference in IC2003, helped interpretation of
the results
Why circulation? Because of insufficient
efficiency of Eq, Difficulty in press. control
for Eq, slow response
Solution! High efficiency of long bubbling
Eq Relief of surface tension effect by 2nd Eq
7
What we learned about the possible error
2 Organic matter (decomposition) gives positive
bias.
Pre-filtration Accumulation in equilibrator
Solution Equilibrator design without filtration
(large inlet or shower head hole). Clean
equilibrator and filter unit (if used) by
frequent maintenance. Large water flow
equilibrator gives relatively small effect even
organic matter decomposes.
8
Day 2 gas pCO2 systems compared with NIES long
Tandem
This equilibrator had filtration unit and
accumulated particulate matter. Filter washed
then Day 3 result was similar to Day 1 result
9
What we learned about the possible error
3 Re-supply air gives positive bias in low pCO2
and negative bias in high pCO2.
Solution Minimize re-supply air. Increase
equilibrator efficiency. Pre-equilibrate the
re-supply air. NDIR in the circulation loop.
Circulating equilibrator needs re-supply air.
Continuous flow circulating system needs large
volume of re-supply air.
10
Day 1 Consistent systems compared with NIES long
Tandem
High bias in low pCO2 and vice versa
11
Day 2 Consistent systems compared with NIES long
Tandem
High bias in low pCO2 and vice versa
12
Day 3 Consistent systems compared with NIES long
Tandem
High bias in low pCO2 and vice versa
13
Idea of pre-equilibration of the re-supply air.
Re -supply air
Air out
Air in
Pre equilibrated air can easily be approximately
equal to the sea water pCO2.
Pre equilibrator/small bubbling equilibrator
14
NDIR in the circulation loop
Small effect of re-supply, however, it is
difficult to keep equilibrator pressure same as
ambient.
15
What we learned about the possible error
4 Equilibrator pressurization or depressurization
Pressurization or depressurization depends on the
location of pump and flow design. Positive or
negative, case by case
Circulation system needs pressure gradient in the
air line.
Solution Precise measurement of equilibrator
pressure Pump design and optimized alignment of
the pump
16
Experiment evidence New Zealand system measured
the equilibrator pressure very precisely
(expensive barometer), and gave appropriate
pressure correction. NRIFS system has pressure
gauge (water height, primitive but very precise)
gave pressure correction. These two systems have
very small (lt1 ppm) bias from NIES Tandem.
17
Two typical circulation equilibrator design
Both design need care about the
pressurization/depressurization. Stop flow
measurement of CO2 does not mean ambient pressure
equilibration. Control of equilibration pressure
to the ambient or measurement of pressure in the
equilibrator is essential.
18
What we learned about the possible error
5 Reactive equilibrator (membrane equilibrator or
chemistry sensors having not inert surface with
CO2) may have memory effect when quick and large
pCO2 change happened.
Our pCO2 change was caused by HCl or NaOH clouds.
The large amplitudes of change and then
stabilized. Memory effect was observed Goatex
equilibrator and chemistry sensors.
19
Product from the WS

• Good laboratory practice (GLP)
• Publication of good on board practice (GOP?) to
  avoid already known possible error