Title: Improving pH System Design and Performance
1Improving pH System Design and Performance
- Roger Reedy - Principal Engineer
- Greg McMillan - Principal Consultant
- John Moulis - Principal Engineer
2Presenters
- Roger Reedy
- Greg McMillan
- John Moulis
3Introduction
- Use of modeling in DeltaV control studio to
optimize design and prototype control of
neutralization process - Topics to be covered
- Existing system
- Project drivers and objectives
- Project cost
- Challenges
- Control system and equipment design
- Process model
- Virtual plant
- Business results
- Summary
4The Luling Plant
5The Luling Plant
6Existing Production System
Filtered Water
Cation
Anion
De-mineralized Water
7Existing Neutralization System
Water
93 Acid
50 Caustic
AT
Cation
Anion
To EO
Final caustic adjustment
Final acid adjustment
Pit
8Project Drivers
- Old pit needs significant upgrades
- Corporate objective for secondary containment
9Project Objectives
- Safe
- Responsible
- Reliable
- Mechanically
- Robust controls, Operator friendly
- Ability to have one tank out of service
- Balance initial capital against reagent cost
10Cost Data
- Acid market price .75/Gal
- Caustic market price .75/Gal
- All other equipment 150k
- Total project cost (2 10k Gal tanks) 550k
2k Gal 5k Gal 10k Gal 20k Gal 40k Gal
Tank 20k 30k 48k 81k 120k
Pump 26k 35k 45k 74k 105k
11Top Ten Signs of a Rough pH Startup
- Food is burning in the operators kitchen
- The only loop mode configured is manual
- An operator puts his fist through the screen
- You trip over a pile of used pH electrodes
- The technicians ask what is a positioner?
- The technicians stick electrodes up your nose
- The environmental engineer is wearing a mask
- The plant manager leaves the country
- Lawyers pull the plugs on the consoles
- The president is on the phone holding for you
12Titration Curve
Slope is 1,000x steeper at 7 pH than at 2 or 12
pH from small buffer effect at 6 and 9 pH (slope
would be 100,000x steeper for pure strong acid
and base)
B Er 100 Fimax
---- Frmax Frmax A
Fimax Er B / A Ss 0.5
Er Where A distance of center of
reagent error band on abscissa from origin B
width of allowable reagent error band on
abscissa for control band Er allowable
reagent error () Frmax maximum reagent flow
(kg per minute) Fimax maximum influent flow (kg
per minute) Ss allowable stick-slip
(resolution limit) ()
Mistake in equipment, piping, valve, and
sensor design can cause the system not only to
fail but to fail miserably
13Challenges
- Process gain changes by factor of 1000x
- Final element rangeability needed is 10001
- Final element resolution requirement is 0.1
- Concentrated reagents (50 caustic and 98
sulfuric) - Caustic valves ¼ inch port may plug at lt 10
position - Must mix 0.05 gal reagent in 5,000 gal lt 2
minutes - Volume between valve and injection must be lt 0.05
gal - 0.04 pH sensor error causes 20 flow feedforward
error - Extreme sport - extreme nonlinearity,
sensitivity, and rangeability of pH demands
extraordinary requirements for mechanical,
piping, and automation system design
14Choices
Really big tank and thousands of mice each with
0.01 gallon of acid or caustic
or
modeling and control
15New Control System
middle signal selector
AY 3-1
Anion
AT 3-2
AT 3-1
AT 3-3
Control Logic CL 1,2,3,4,5,6,7,8 optimizes
system operation
Cation
FT 3-4
Static mixer
Influent
Signal characterizers provide Linear Reagent
Demand Control
signal characterizer
signal characterizer
AY 2-3
LT 2-5
AY 1-3
LC 2-5
LT 1-5
LC 1-5
pH set point
pH set point
signal characterizer
signal characterizer
CL2
CL1
AY 2-2
AC 2-1
AY 1-2
AC 1-1
middle signal selector
middle signal selector
CL7
CL7
splitter
splitter
Tank 1
Tank 2
AY 2-4
AY 2-1
AY 1-1
AY 1-4
NaOH
Acid
NaOH
Acid
AT 2-3
AT 2-2
AT 2-1
AT 1-2
AT 1-1
FT 2-1
AT 1-3
FT 2-2
FT 1-1
FT 1-2
eductors
eductors
CL3
CL4
CL6
CL6
CL8
CL8
CL5
16Control Logic
- If influent pH more than 2 pH units away from
control band and flow has just started, head
start tank 1 pH loop for 2 minutes to move 2 pH
closer (CL1) - If Tank 1 pH more than 2 pH units away from
control band and flow has just started, head
start tank 2 pH loop for 2 minutes to move 2 pH
closer (CL2) - If Tank 1 pH within control band, reduce its
level rapidly to minimum (CL3) - If Tank 2 pH within control band, reduce its
level rapidly to minimum (CL4) - If Tank 2 pH outside control band, and level in
Tank 2 is higher than Tank 1, level control
recirculation of Tank 2 back to Tank 1 (CL5) - If caustic reagent valve signal is less than 10,
use pulse width modulation and increase pH loop
filter time and reset time to smooth out pulses
(CL6) - Shut reagent valves periodically for 30 seconds
to get tank pH reading to optimize the
recirculation pH set point (CL7) - If feed is negligible and tank pH within control
band, shut off pump (CL8)
17Middle Signal Selection Advantages
- Inherently ignores single measurement failure of
any type including the most insidious PV failure
at set point - Inherently ignores slowest electrode
- Reduces noise and spikes particularly for steep
curves - Offers online diagnostics on electrode problems
- Slow response indicates coated measurement
electrode - Shortened response indicates aged measurement
electrode - Drift indicates coated or contaminated reference
electrode - Noise indicates dehydrated measurement electrode
- Facilitates online calibration of a measurement
18Linear Reagent Demand Control
- Signal characterizer translates PV and SP from pH
to Reagent Demand - PV is abscissa of the titration curve scaled 0 to
100 reagent demand - Piecewise segment fit normally used to go from
ordinate to abscissa of curve - Fieldbus block offers 21 custom space X,Y pairs
(X is pH and Y is demand) - Closer spacing of X,Y pairs in control region
provides needed compensation - Special configuration is needed to provide
operations with pH interface to - See loop PV in pH and enter loop SP in pH
- Set point on steep part of curve shows biggest
improvements from - Reduction in limit cycle amplitude seen from pH
nonlinearity - Decrease in limit cycle frequency from final
element resolution (e.g. stick-slip) - Decrease in crossing of split range point
- Reduced reaction to measurement noise
- Shorter startup time (loop sees real distance to
set point and is not detuned) - Simplified tuning (process gain no longer depends
upon titration curve slope) - Restored process time constant (slower pH
excursion from disturbance)
19Dynamic Process Model in DeltaV
Streams, pumps, valves, sensors, tanks, and
mixers are modules from DeltaV composite
template library.
Each wire is a pipe that is a process stream
data array (e.g. pressure, flow, temperature,
density, heat capacity, and concentrations)
First principle conservation of material,
energy, components, and ion charges
20DeltaV Virtual Plant
DCS batch and loop configuration, displays, and
historian
Embedded Advanced Control Tools
Embedded Modeling Tools
Dynamic Process Model
Loop Monitoring And Tuning
Virtual Plant Laptop or Desktop Personal
Computer Or DCS Application Station or Controller
Online Data Analytics
Model Predictive Control
Process Knowledge
21Top Ten Reasons We Use a Virtual Plant
- You cant freeze, restore, and replay an actual
plant batch - No software to learn, install, interface, and
support - No waiting on lab analysis
- No raw materials
- No environmental waste
- Virtual instead of actual problems
- Batches are done in 5 minutes instead of 5 hours
- Plant can be operated on a tropical beach
- Last time we checked our wallet we didnt have
1,000K - Actual plant doesnt fit in our suitcase
22Business Results Achieved
- Results of modeling indicate all objectives will
be met - Rework savings on 550k project potentially
significant - Equipment savings of 132k per tank (10k vs. 40k
Gal)
23Summary
- Study shows potential project savings overwhelm
reagent savings - Reagent savings is lt 1K per year
- Equipment savings is 132K per tank
- Modeling removes uncertainty from design
- First principle relationships show how well
mechanical, piping, and automation system deal
with nonlinearity, sensitivity, and rangeability - Modeling enables prototyping of control
improvements - Linear reagent demand control speeds up response
from PV on flat and oscillations from PV on steep
part of titration curve - Control logic optimizes pH loops to minimize
inventory to maximize availability and turn off
pumps to reduce energy use - Initialization of pH loop when feed flow starts
provides head start for upset - Pulse width modulation of caustic at low valve
positions minimizes plugging - Periodic optimization of loop set point keeps
tank pH within control band - Feedback?
- Questions?
24Where To Get More Information
- McMillan, Gregory and Cameron, Robert, Advanced
pH Measurement and Control, 3rd edition, ISA,
2005 - McMillan, Gregory K., A Funny Thing Happened on
the Way to the Control Room, 1989
http//www.easydeltav.com/controlinsights/FunnyThi
ng/default.asp - McMillan, Gregory K., Plant Design and Education
Categories, http//ModelingandControl.com - McMillan, Gregory, K. and Sowell, Mark. S.,
Virtual Control of Real pH, Control, November
2007 - McMillan, Gregory, K. and Sowell, Mark. S.,
Advances in pH Modeling and Control, ISA 54th
IIS Paper IIS08-P044, 2008