Automation Development in Water and Wastewater Systems

Abstract

Advanced control is getting increasingly demanded in water and wastewater treatment systems. Various case studies have shown significant savings in operating costs, including energy costs, and remarkably short payback times. It has been demonstrated that instrumentation, control and automation (ICA) may increase the capacity of biological nutrient removing wastewater treatment plants by 10-30% today. With further understanding and exploitation of the mechanisms involved in biological nutrient removal the improvements due to ICA may reach another 20-50% of the total system investments within the next 10-20 years. Disturbances are the reason for control of any system. In a wastewater treatment system they are mostly related to the load variations, but many disturbances are created also within the plant. In water supply systems some of the major disturbances are related the customer demand as well as to leakages or bursts in the pipelines or the distribution networks. Hardly any system operates in steady state but is more or less in a transient state all the time. Water and energy are closely related. The role of energy in water and wastewater operations is discussed. With increasing energy costs and the threatening climate changes this issue will grow in importance.

Full text

Environ. Eng. Res. Vol. 12, No. 5, pp. 197
～
200, 2007
Korean Society of Environmental Engineers
Automation Development in Water and Wastewater Systems
Gustaf Olsson
†
Department of Industrial Electrical Engineering and Automation, Lund University, Lund, Sweden
Received November 2007, accepted December 2007
Abstract
Advanced control is getting increasingly demanded in water and wastewater treatment systems. Various case studies have shown significant
savings in operating costs, including energy costs, and remarkably short payback times. It has been demonstrated that instrumentation, control and
automation (ICA) may increase the capacity of biological nutrient removing wastewater treatment plants by 10-30% today. With further
understanding and exploitation of the mechanisms involved in biological nutrient removal the improvements due to ICA may reach another 20-50%
of the total system investments within the next 10-20 years. Disturbances are the reason for control of any system. In a wastewater treatment system
they are mostly related to the load variations, but many disturbances are created also within the plant. In water supply systems some of the major
disturbances are related the customer demand as well as to leakages or bursts in the pipelines or the distribution networks. Hardly any system
operates in steady state but is more or less in a transient state all the time. Water and energy are closely related. The role of energy in water and
wastewater operations is discussed. With increasing energy costs and the threatening climate changes this issue will grow in importance.
Keywords: Instrumentation, control and automation (ICA)
1. Driving Forces and Motivations for Control
1
Instrumentation, control and automation (ICA) got the atten-
tion in the water and wastewater industry already in the 1970s.
Still, however, dynamical systems and process control is seldom
part of the general civil engineering or environmental engineer-
ing curricula. Consequently many water and wastewater sys-
tems designers are unaware of the potential of ICA. It has been
demonstrated that ICA may increase the capacity of biological
nutrient removal (BNR) wastewater treatment plants (WWTP)
by 10-30% today. The advanced knowledge of the mechanisms
involved in biological nutrient removal that is being gained
today is producing an increased understanding of the processes
and the possibility to control them. There is a sophisticated
relationship between the operational parameters in a treatment
system and its microbial population and biochemical reactions,
and hence its performance. With further understanding and
exploitation of these relationships the improvements due to ICA
may reach another 20-50% of the total system investments
within the next 10-20 years. Various case studies of advanced
control in water and wastewater treatment systems have shown
significant savings in operating costs and remarkably short
†
Corresponding author
E-mail: Gustaf.olsson@iea.lth.se
Tel: +46-46-222-4788, Fax: +46-46-14-21-14
payback times (Olsson et al., 2005).
2. Disturbances
A major incentive for control is the presence of disturbances,
and the impact of them has to be compensated. Compared to
most other process industries, the disturbances that a wastewater
treatment plant is subject to are extremely large. The wastewater
influent typically varies substantially both in its concentration,
composition and flow rate, with time scales ranging from frac-
tion of hours to months. Discrete events such as rainstorms,
toxic spills and peak loads may also occur from time to time.
As a result, the plant is hardly ever in steady state, but is subject
to transient behavior all the time (Olsson & Newell, 1999).
In a water supply system the major disturbances instead occur
at the load or customer side. The inflow to the water plant can
be kept relatively constant, while the demand from the customers
will vary depending on the time of the day, the weather and the
season. Sudden bursts or leakages cause disturbances than look
more like discrete events. They have to be discovered at an
early stage and countermeasures - preferably automatic - have
to be realized quickly.
Consistent performance must be maintained despite the distur-
bances. The traditional way of dampening the disturbances has
been to design plants with large volumes to attenuate large load

Gustaf Olsson
198
disturbances. This solution incurs large capital costs. On-line
control systems, which have been demonstrated to cope well
with most of these variations, are a much more cost-effective
and thus attractive alternative. Disturbance rejection is indeed
one of the major incentives for introducing on-line process
control. Many disturbances in a wastewater treatment system
are related to the plant influent flow. Any of these changes have
to be measured and compensated for. If the effect of the
disturbance is measured within the plant, such as a change in
the dissolved oxygen level, a rising sludge blanket, or a varying
suspended solids concentration, the measured information is fed
back to a controller that will activate a pump, a valve, or a com-
pressor, so that the influence on the plant behavior is minimized.
Too often unnecessary disturbances are created within the
plant itself. Often this depends on a lack of understanding how
the various parts of the plant interact. Just one example: if the
influent flow rate cannot be varied continuously but the pumps
are operated in an on/off mode the consequence is that the plant
will be subject to sudden flow rate changes. In particular, the
settler operation will suffer from such sudden flow rate changes.
Recycling of water and sludge in a wastewater treatment
plant creates apparent couplings between various unit processes.
If these interactions are not considered, then the plant operation
will suffer. For example, if sludge supernatant is recycled to the
plant influent during a high load, then the nitrogen load to the
plant may be very large and can be measured as an increase in
the oxygen uptake rate. It is crucial to identify the sources of
disturbances in order to obtain a high performance operation of
a plant. Then the control system can be structured so that distur-
bances are attenuated or even avoided (Olsson et al., 2005).
Further internal disturbances may be generated due to inade-
quate or inappropriate operations including human errors,
unsuitable or malfunctioning actuators and/or sensor break-
downs. These may potentially cause major operational problems.
Many of the internal disturbances may be avoided (or their
impacts minimized) through introducing on-line control sys-
tems, including early warning systems.
3. The Role of Control and Automation
ICA in wastewater treatment systems have come a long way
and is now an established and recognized area of technology in
the profession. A number of factors have combined made this
progress possible:
∙Instrumentation technology is today so much more mature.
Complex instruments like on-line in-situ nutrient sensors
and respirometers are now regularly used in the field.
∙Actuators have improved over the years. Today variable
speed drives in pumps and compressors are commonly
used to allow a better controllability of the plant;
∙Computing power can be considered almost “free”;
∙Data collection is no longer a great obstacle. Software
packages and SCADA systems are available for data
acquisition and plant supervision;
∙Control theory and automation technology offer powerful
tools. Benchmarking and various tools for evaluating control
strategy performance have been developed;
∙Advanced dynamical models of many unit processes have
been developed. Commercial simulators are available to
condense the knowledge of plant dynamics;
∙Operators and process engineers are often educated in inst-
rumentation, computers and control ideas. However, there
is still a great need for better education in these areas.
∙There are obvious incentives for ICA, not the least from an
economic point of view. Plants are also becoming increa-
singly complex which necessitates automation and control.
Today the main obstacle for more ICA is the lack of process
flexibility. Plant design and operation still have to be integrated
in a systematic way.
4. Instrumentation and Monitoring
To measure is to know. Developments during the last two
decades have contributed that instrumentation is not the main
obstacle for ICA (Olsson & Newell, 1999; Vanrolleghem &
Lee, 2003; Olsson et al., 2005). The increased confidence in
instrumentation is now driven by the fact that clear definitions
of performance characteristics and standardized tests for inst-
rumentation have become available (ISO 15839:2003).
To track the process operational state via the instrumentation
is called monitoring. For the clean water supply on-line monito-
ring will be required throughout the system including at the tap.
The availability of low cost instrumentation will encourage better
leakage detection and water quality monitoring. In wastewater
treatment systems the use of ICA has proven to significantly
reduce the costs for operation. However, even reliable instru-
mentation can fail during operation, which can have serious
consequences if the instrumentation is used in closed loop con-
trol. Therefore real time data validation is needed before using
measurements for control purposes (Lynggaard-Jensen & Frey,
2002). If confidence in a measurement decreases, it might be
possible (on a short-term basis) to use an estimated value, but
eventually control must be set to a default scheme until confi-
dence in the measurement has been restored.
In a sophisticated treatment plant there is a huge data flow
from the process. More instrumentation will further provide
more data. Unlike humans, computers are infinitely attentive
and can detect abnormal patterns in plant data. The capability
of computers to extract patterns (useful information) is rarely
utilized beyond simple graphing. Information technology is not
commonly used to encapsulate process knowledge, i.e. know-
ledge about how the process works and how to best operate it.
Process knowledge is typically built up from the experience of
operators and engineers but all too often disappears with them
when they leave. If process knowledge can be encapsulated,
then not only is it retained but the computer can also assist
decision-making in plant operation (Rosen et al., 2004). The
potential of substantial operator support for diagnosis and for
corrective actions is there and has been demonstrated, but it
needs to be adopted by the water and wastewater industry.
5. Control Applications in Wastewater Treatment

Automation Development in Water and Wastewater Systems
199
The fundamental principle of control is feedback. The process
(for example, an aerator, a chemical dosage system, or an anae-
robic reactor) is all the time subject to disturbances. The current
state of the process has to be measured by some sensor and this
is the basis for a decision. In order to make a decision the goal
or purpose of has to be expressed. Having made the decision it
has to be implemented via an actuator, which is typically a motor,
a pump, a valve or a compressor. In other words: control is
about how to operate the plant or process towards a defined
goal, despite disturbances (Olsson & Newell, 1999).
The traditional WWTP control is still unit process oriented to
a great extent. Some examples of state-of-the-art control are
mentioned here (Olsson et al., 2005) :
∙DO control with a constant or a variable setpoint as part of
the aerator unit process operation;
∙Aeration phase length control in alternating plants is based
on nutrient sensors, but still locally;
∙Nitrate recirculation control in a pre-denitrification plant
can be based on nitrate and DO measurements in the
aerator and in the anoxic zone (Ingildsen, 2002);
∙Advanced sludge retention time control is based on local
measurements of effluent ammonia concentration and of
estimates of nitrification capacity;
∙Return sludge control can be based on sludge blanket
measurements in the settler;
∙Aeration tank settling (ATS) is one way of temporarily
increasing the plant capacity at storm conditions (Nielsen
et al., 1996, Gernaey et al., 2004);
∙The control of anaerobic processes aims at regulating the
biogas flow, at stabilizing the process and at maximizing
its productivity. Still current state-of-the-art focuses on the
unit process operation;
∙Successful chemical precipitation control can be based on
local measurements of phosphate concentration.
In water supply systems leakage detection has been success-
fully applied for many different operating conditions. Leakages
can appear as sudden bursts or slow and gradual leakages. They
will take place in both single water transmission lines and in
water distribution networks. Many interesting methods have
been developed to cope with leakages, and some interesting
examples are shown in Misiunas (2005a, 2005b) that contain
further references.
6. Energy and Water
Energy and water are closely related, which is seldom consi-
dered. Here we will briefly discuss the consumption of electri-
cal energy and the production of biogas energy.
6.1. Electrical Energy
Treatment and transmission of water and wastewater requires
large amounts of energy. In a country like Sweden water and
wastewater operations use about 1% of the total national electri-
cal energy supply. The demand on electrical energy will have
an environmental impact, which means that the sustainability
issue is critical also from an energy perspective. Clean water
requires electrical energy; for pumping of drinking water and of
sewage, for mixing and for aeration of wastewater, for chemicals,
and for transportation of sludge. Desalination for water supply
is rapidly increasing. In the Mediterranean area there is an 18%
annual increase and in Saudi Arabia a 17% increase every year.
Impressive efforts are in place in Korea. This just demonstrates
that the energy issue will require a lot more attention.
As long as the cost of electrical energy has been quite low
the energy aspect has not been given much attention. However,
as prices are raising the interest in various energy savings has
been increasing. Many different assessments can be defined for
energy requirement, such as kWh/person/year or kWh/kg N
removed etc. Here we will not elaborate on various methods to
estimate the energy use. Instead we will point at some impor-
tant factors where control and automation can bring down the
electrical energy requirement.
Dissolved oxygen control will save a lot of electrical energy
compared to no control at all. A time varying setpoint of the
DO concentration will further reduce the energy consumption
(Olsson & Newell 1999, Olsson et al., 2005). Large pumps,
primarily for the influent water, are often the most energy de-
manding equipment in a plant. Too often the pumping equip-
ment has not been designed for the adequate flow rates. Aera-
tion by compressors ought to be continuously variable. To
control airflow by closing airflow valves will cause a lot of
energy losses. Instead, variable speed compressors will save
energy significantly.
A wastewater treatment plant in fact should be considered a
recovery plant for both nutrients and energy. If we consider the
energy potential in anaerobic digestion there is a huge unused
potential in most places. We can illustrate this with one good
example (the Rya WWTP in Göteborg, Sweden): the plant uses
41 kWh/person/year of electrical energy. At the same time the
plant produces biogas corresponding to 72 kWh/person/year.
The heat content of the effluent water is taken care of in heat
pumps. Here the production potential is 336 kWh/person/year.
The plant is in fact an important energy producer.
6.2. Biogas Production
Recent data show that anaerobic digestion (AD) uses only
some 20% of the energy content of the sewage. In addition,
costs of sludge transportation and disposal, which currently
place a major burden on the industry, could be reduced. The
nature of the influent characteristics involves dynamic variation
in both flow rate and composition (Batstone et al., 2002). Hand-
ling of these disturbances by attenuation or rejection is thus
important for stable operation. Many anaerobic bioreactors are
still being operated without close monitoring and control. This
is not only due to the fact that the anaerobic process involves a
complicated mechanism of degradation steps, but it is also due
to the lack of proper analytical devices. In fact, sensor techno-
logy is the weakest part of the process chain (Liu, 2003; Olsson
et al., 2005).
Close monitoring and control makes it possible to enhance

Gustaf Olsson
200
the operational stability, to attenuate and reject disturbances
and to allow the treatment of waste and biogas production at a
higher specific rate (Liu et al., 2004). The activity of the differ-
ent microbial groups involved in the AD process can be mea-
sured indirectly by monitoring the metabolites. In general, it is
now possible to analyse pH, alkalinity, biogas flow and com-
position, VFAs, biodegradable organic matter, dissolved hydro-
gen, and toxicity on-line by less expensive sensors and instru-
ments. Usually the feed rate is the control variable. Another
interesting approach reported in recent years is the probing
control strategy based on analysing the effect of disturbances
added on purpose to the influent flow rate (Steyer et al., 1999).
7. Concluding Remarks
Disturbances are everywhere and are the main reason for
control. Uncertainty in the process or in its environment makes
automation both an opportunity and a great challenge. Applica-
tion of automation in water operations can be said to have two
primary functions: information acquisition and process control.
For the former function, the level of automation is relatively
high. Often many thousands of variables are gathered on-line in
the SCADA systems of treatment plants and more or less sophi-
sticated data analyses are standard components of the treatment
operation. The latter function, process control, is less developed
and often limited to a few unit process control loops. Future
development will be exploiting the enormous capacity of data
distribution that is possible today. Many SCADA systems are
also applying the technology from the Internet, which gives an
almost unlimited potential for remote data evaluation and deci-
sion. The distributed control room is already here. There is a
limit of how much expertise a treatment plant can afford. How-
ever, given that plant data can be made available anywhere it is
possible to utilize specialist competencies wherever they are
located.
The increasing incorporation of ICA in water treatment opera-
tion is not only driven by the impressive technical development
of instrumentation and computer technology, modelling and
control, and the progress in automation. It is motivated by eco-
nomy and environmental obligations and turns out to be a nece-
ssary and worthwhile investment. It is already proven in several
installations that ICA investments have paid off quickly and we
will see that ICA will become an increasing part of the total
investments.
References
1. Batstone, D. J., Keller, J., Angelidaki, R. I., Kalyuzhnyi, S.
V., Pavlostathis, S. G., Rozzi, A., Sanders, W. T. M.,
Siegrist, H., and Vavilin, V. A., “Anaerobic Digestion Model
No.1.” Scientific and Technical Report No.13, IWA
Publishing, London (2002).
2. Gernaey, K. V., Nielsen, M. K., Thornberg, D., Höök, B.,
Munk-Nielsen, T., Ingildsen, P., and Jørgensen, S. B., “First
principle suspended solids distribution modelling to sup-
port ATS introduction on a recirculating WWTP,” 2
nd
Int.
IWA Conf. on Automation in Wat. Quality Monitoring,
Vienna (2004).
3. Ingildsen, P., Realising Full-Scale Control in Wastewater
Treatment Systems Using in Situ Nutrient Sensors, PhD
thesis, Dept. of Ind. Electrical Engineering and Automation
(IEA), Lund Univ., Lund, Sweden, Available at www.iea.
lth.se/publications (2002).
4. Water Quality - On-line sensors/analysing equipment for
water - Specifications and performance tests, First edition,
ISO15839 (2003).
5. Liu, J., Instrumentation, Control and Automation in Anae-
robic Digestion, PhD thesis, Department of Biotechnology,
Lund University, Sweden, ISBN: 91-89627-17-2 (2003).
6. Liu, J., Olsson, G., and Mattiasson, B., “Control of an Anae-
robic Reactor towards Maximum Biogas Production,” Wat.
Sci. Tech., 50(11), 189-198 (2004).
7. Lynggaard-Jensen, A., and Frey, M., Use and Reporting of
Data - Data Handling and Validation, In On-line Monito-
ring for Drinking Water Utilities, American Water Works
Association Research Foundation, Denver, ISBN 1-58321-
183-7 (2002).
8. Misiunas, D., Lambert, M. F., Simpson, A. R., and Gustaf
Olsson, “Burst detection and Location in water distribution
networks,” Wat. Sci. Tech. / Water Supply, 5(3-4) (2005a).
9. Misiunas, D., Vitkovsky, J., Olsson, G., Simpson, A. R.,
and Lambert, M. F., “Pipeline Break Detection Using the
Transient Monitoring,” Journal of Water Resources Plan-
ning and Management, ASCE, 131(4), 316-325 (2005b).
10. Nielsen, M. K., Carstensen, J., and Harremöes, P., “Combined
control of sewer and treatment plant during rainstorm,”
Wat. Sci. Tech., 34(3-4), 181-187 (1996).
11. Olsson, G., and Newell, B., “Wastewater Treatment Systems,
Modelling, Diagnosis and Control,” IWA Publishing,
London (1999).
12. Olsson, G., Nielsen, M. K., Yuan, Z., Lynggaard-Jensen,
A., and Steyer, J. P., “Instrumentation, Control and Auto-
mation in Wastewater Treatment Systems,” Scientific and
Technical Report No.15, IWA Publishing, London (2005).
13. Rosen, C., Ingildsen, P., Guildal, T., Nielsen, M. K., Jacob-
sen, B. N., and Ønnerth, T., “On-line estimation of nitrifica-
tion and denitrification capacity at Avedøre wastewater
treatment plant,” Proc. Int. Conf. Upgrading of Wastewater
Treatment Plants (AquaTech2004), Amsterdam (2004).
14. Steyer, J. P., Buffiére, P., Rolland, D., and Moletta, R.,
“Advanced control of anaerobic digestion processes through
disturbances monitoring,” Wat. Res., 33(9), 2059-2068
(1999).
15. Vanrolleghem, P. A., and Lee, D. S., “On-line monitoring
equipment for wastewater treatment processes: State of the
art,” Wat. Sci. Tech., 47(2), 1-34 (2003).