WASTE WATER TREATMENT PLANT

A wastewater treatment plant is a facility that processes wastewater from homes, businesses,
and industries to remove impurities before it is released into the environment 1. The treatment
process involves physical, chemical, and biological processes to remove contaminants such as
organic matter, nutrients, and pathogens 12. The treated water can then be safely discharged into
rivers, lakes, or oceans, or used for irrigation or other purposes 1.
There are several types of wastewater treatment plants, including sewage treatment
plants 2. The history of wastewater treatment dates back to ancient times, with many ancient
cities having drainage systems to carry rainwater away from roofs and pavements 1. The first
sewage treatment plant was built in London in the 19th.

Wastewater treatment plants (WWTP) are crucial in maintaining water quality and resource
recovery in a world facing growing challenges such as climate change, water-energy-food nexus
and the increase of legal requirements and users’ expectations1. The efficiency of a WWTP is
determined by the amount of pollutants removed from the wastewater and the amount of energy
consumed during the treatment process2.
There are several ways to improve the efficiency of a WWTP. One way is to adopt advanced
treatment technologies such as Membrane Bioreactor (MBR), Ultrafiltration (UF), and Reverse
Osmosis (RO). Studies have shown that MBR technology can reduce energy consumption by up
to 50% compared to traditional wastewater treatment methods2. Another way to improve
efficiency is to optimize the operation of the plant. This can be achieved by adopting a
performance assessment system (PAS) which assesses operating conditions and identifies critical
aspects of the WWTP which can negatively affect its effectiveness,

Stage One — Bar Screening

Removal of large items from the influent to prevent damage to the facility’s pumps, valves and other
equipment.
The process of treating and reclaiming water from wastewater (any water that has been used in homes, such as
flushing toilets, washing dishes, or bathing, and some water from industrial use and storm sewers) starts with
the expectation that after it is treated it will be clean enough to reenter the environment.

The quality of the water is dictated by the Environmental Protection Agency (EPA) and the Clean Water Act,
and wastewater facilities operate to specified permits by National Pollutant Discharge Elimination System
(NPDES). According to the EPA, The Clean Water Act (CWA) establishes the basic structure for regulating
discharges of pollutants into the waters of the United States and regulating quality standards for surface waters.
Under the CWA, EPA sets wastewater standards for industry. The EPA has also developed national water
quality criteria recommendations for pollutants in surface waters. EPA's National Pollutant Discharge
Elimination System (NPDES) permit program controls discharges.
As an example of expected standards, the Biochemical Oxygen Demand (BOD) of average wastewater effluent
is 200 mg/L and the effluent after treatment is expected to be >30 mg/L. It is crucial a wastewater facility
meets these expectations or risk stiff penalty.

The physical process of wastewater treatment begins with screening out large items that have found their way
into the sewer system, and if not removed, can damage pumps and impede water flow. A bar screen is usually
used to remove large items from the influent and ultimately taken to a landfill.

Related White Papers

Water Quality and the Clean Water Rule

How to Select Pocket pH Meter for Applications

Stage Two — Screening

Removal of grit by flowing the influent over/through a grit chamber.
Fine grit that finds its way into the influent needs to be removed to prevent the damage of pumps and
equipment downstream (or impact water flow). Too small to be screened out, this grit needs to be removed
from the grit chamber. There are several types of grit chambers (horizontal, aerated or vortex) which control
the flow of water, allowing the heavier grit to fall to the bottom of the chamber; the water and organic material
continue to flow to the next stage in the process. The grit is physically removed from the bottom of the
chamber and discarded.

Stage Three — Primary Clarifier

Initial separation of solid organic matter from wastewater.
Solids known as organics/sludge sink to the bottom of the tank and are pumped to a sludge digestor or sludge
processing area, dried and hauled away. Proper settling rates are a key indicator for how well the clarifier is
operating. Adjusting flow rate into the clarifier can help the operator adjust the settling rates and efficiency.

After grit removal, the influent enters large primary clarifiers that separate out between 25% and 50% of the
solids in the influent. These large clarifiers (75 feet in diameter, 7½ inches at the edges and 10½ feet in the
center as an example) allow for the heavy solids to sink to the bottom and the cleaner influent to flow. The
effectiveness of the primary clarification is a matter of appropriate water flow. If the water flow is too fast, the
solids don’t have time to sink to the bottom resulting in negative impact on water quality downstream. If the
water flow is too slow, it impacts the process up stream.

The solids that fall to the bottom of the clarifier are know as sludge and pumped out regularly to ensure it
doesn’t impact the process of separation. The sludge is then discarded after any water is removed and
commonly used as fertilizer.

Stage Four — Aeration

Air is pumped into the aeration tank/basin to encourage conversion of NH3 to NO3 and provide oxygen for
bacteria to continue to propagate and grow.
Once converted to NO3, the bacteria remove/strip oxygen molecules from the nitrate molecules and the
nitrogen (N) is given off as N2↑ (nitrogen gas).

At the heart of the wastewater treatment process is the encouragement and acceleration of the natural process
of bacteria, breaking down organic material. This begins in the aeration tank. The primary function of the
aeration tank is to pump oxygen into the tank to encourage the breakdown of any organic material (and the
growth of the bacteria), as well as ensure there is enough time for the organic material to be broken
down. Aeration can be accomplished with pumping and defusing air into the tank or through aggressive
agitation that adds air to the water. This process is managed to offer the best conditions for bacterial growth.
Oxygen gas [O2] levels below 2 ppm will kill off the bacteria, reducing efficiency of the plant. Dissolved
oxygen monitoring at this stage of the plant is critical. Ammonia and nitrate measurements are common to
measure how efficient the bacteria are in converting NH3 to N2↑.

A key parameter to measure in wastewater treatment is Biochemical Oxygen Demand (BOD). BOD is a
surrogate indicator for the amount of organic material present and is used to determine the effectiveness of
organic material breakdown. There are a number of other tests used to ensure optimal organic material
breakdown (and BOD reduction) such as measuring pH, temperature, Dissolved Oxygen (DO), Total
Suspended Solids (TSS), Hydraulic Retention Time (flow rate), Solids Retention Time (amount of time the
bacteria is in the aeration chamber) and Mixed Liquor Suspended Solids. Ongoing and accurate monitoring is
crucial to ensure the final required effluent BOD.

Stage Five — Secondary Clarifier

Treated wastewater is pumped into a secondary clarifier to allow any remaining organic sediment to settle out
of treated water flow.
As the influent exits the aeration process, it flows into a secondary clarifier where, like the primary clarifier,
any very small solids (or fines) sink to the bottom of the tank. These small solids are called activated sludge
and consist mostly of active bacteria. Part of this activated sludge is returned to the aeration tank to increase
the bacterial concentration, help in propagation, and accelerate the breakdown of organic material. The excess
is discarded.

The water that flows from the secondary clarifier has substantially reduced organic material and should be
approaching expected effluent specifications.
Stage Six — Chlorination (Disinfection)
Chlorine is added to kill any remaining bacteria in the contact chamber.
With the enhanced concentration of bacteria as part of the aeration stage, there is a need to test the outgoing
effluent for bacteria presence or absence and to disinfect the water. This ensures that higher than specified
concentrations of bacteria are not released into the environment. Chlorination is the most common and
inexpensive type of disinfection but ozone and UV disinfection are also increasing in popularity. If chorine is
used, it is important to test for free-chlorine levels to ensure they are acceptable levels before being released
into the environment.

Stage Seven — Water Analysis & Testing

Testing for proper pH level, ammonia, nitrates, phosphates, dissolved oxygen, and residual chlorine levels to
conform to the plant’s NPDES permit are critical to the plant’s performance.
Although testing is continuous throughout the wastewater treatment process to ensure optimal water flow,
clarification and aeration, final testing is done to make sure the effluent leaving the plant meets permit
specifications. Plants that don`t meet permit discharge levels are subject to fines and possible incarceration of
the operator in charge.

Stage Eight — Effluent Disposal

After meeting all permit specifications, clean water is reintroduced into the environment.
Although testing is continuous throughout the wastewater treatment process to ensure optimal water flow,
clarification and aeration, final testing is done to make sure the effluent leaving the plant meets permit
specifications. Plants that don`t meet permit discharge levels are subject to fines and possible incarceration of
the operator in charge.
Due to the growth in the human population globally, it is noted that various industries have also grown. The need for
an excess supply of water and the generation of high effluent quality upon proper treatment technologies has become
a necessity. These two crucial needs can be achieved with the aid of membrane bioreactor (MBR) that has been
proven to be effective in removing organic and inorganic matters as a biological unit for wastewater treatment. MBR
plants are created by integrating the biological process with membrane filtration which possesses numerous benefits
if compared with conventional methods such as activated sludge; MBR is widely used for municipal and industrial
wastewater treatment. This review addresses basic concepts of MBRs plants and subsequently provides information
on the recent developments of each part related to MBR plants. The characteristics of the bioreactor treatment
process is discussed in detail, and then a comprehensive review of the membrane separation process is examined. The
fouling phenomena as a main obstacle to widespread MBRs plant is presented in detail with recent fouling mitigation
methods. The efforts of a number of novel MBR processes are summarized. In order to tackle the existing limitation
of MBRs to be practical on a larger scale, the existing challenges and future research efforts are proposed

2. Biological wastewater treatment

Generally, the bioreactor operating conditions highly influence such characteristics of the microorganism as size,
content of filamentous microorganisms, growth rate, etc. On the other hand, the activity of microorganisms can affect
the performance of the MBR in two different ways; in the quality of the effluent and how much the MBR can treat the
wastewater pollutant, and the fouling properties of the membranes. Hence, profound study of the principle of the
biological wastewater treatment such as microbiology, metabolism of microorganism, microbial stoichiometry, and
kinetics in bioreactor is necessary in order to determine the optimum operating conditions of the bioreactor and
design characteristics of the MBR plants [9].
The structure and composition of the microbial community of a bioreactor varies from one MBR plant to another and
across time scale for a given MBR unit. The main reason for this variation is the important features of the
microorganisms in environmental engineering systems, including MBR plants. Due the influence of wastewater from
the atmosphere, which is fed into the bioreaction, diverse microorganisms are structured into variety of communities.
However, by adjusting operating conditions and reactor design, specific type of microorganisms can be enriched in
the bioreactor [10].
The type of microorganism and their functionality are the same in both CAS and MBR plants. However, their
characteristics are different due to long SRT and high concentration of biomass in MBR bioreactors which results in
Ref. [11]:
 1.
 maintaining slow-growing microorganisms compared to shorter SRTs of CAS aeration
tank which is beneficial due to degradation of the recalcitrant organic matters.
Nevertheless, it can also generate unwanted microorganisms such as foaming
microorganisms, and

 2.
reducing the fraction of active biomass from total solids in the bioreactor by producing
more inert solids.

ype of microorganisms
In the bioreactor, five major groups of microorganisms are generally found: Bacteria
(e.g., Proteobacteria), Protozoa (e.g., Amoebae, Flagellates, Ciliates), Metazoa (e.g., Rotifers,
Nematodes, Tartigrades), filamentous bacteria, algae, and fungi [12]. However, the majority of
microorganisms (over 90%) that exist in the activated sludge are bacteria [12]. Most of the
bacteria accumulate together and form pairs, chains, or clusters but they can also continue their
life while living as a single cell. Regarding their metabolisms, they can use numerous sources of
energy, electron donors, electron acceptors, and carbon sources. Their adaptability can be used
helpfully not only to treat different types of organic and inorganic pollutants but also provides
conditions to treat specific type of materials that may exist in the wastewater. Shchegolkova
et al. [13] found the structural of bacteria communities in activated sludge and incoming sewage
for three different wastewater treatment plants by performing 16S rRNA gene sequencing.
Additionally, they provide a heatmap includes the top 40 bacteria families in AS (i.e., 94.2-
97.5% of all bacteria) [13].
Tendency of microorganism to accumulate onto the surface of membrane causes creation of
biofilms. With the aid of self-produced matrix material in contrary to the planktonic
cells (i.e., extracellular polymeric substance (EPS)) biofilm cells are embedded. Adhesive
properties of biofilm are mainly due to the presence of proteins and carbohydrates that exist in
the EPS which is an important disadvantage of MBR plants [14].

2.2. Microbial stoichiometry and kinetics in bioreactor

Balanced microbial stoichiometric equations are like the chemical stoichiometric equations while
they are quite important for estimating biological performance and treatments. However, in the
microbial kinetics equation consumed substrate serves as energy source and used for biomass
synthesis simultaneously. In other word, microorganisms are not used only as catalysts for
biological reaction but also through microbial growth during treatment process they will
reproduce themselves as well. The ratio of generated biomass to the consumed substrate (e.g.,
glucose) can be referred as biomass yield or growth which is depended on microbial
composition and growth conditions [13].
With the aid of balanced microbial stoichiometric equations, understanding the elements that are
involved in the reaction (i.e., primary electron donor, terminal electron acceptor, nutrients,
biomass, and oxidized products) and the rate of consumption or production of elements are
important. Nevertheless, prediction of how fast the reaction taking place is impossible.
Determination of the microbial reaction rate is important in order to estimate the required volume
of bioreactor as well as biomass concentration to achieve a specific result. Also, with the aid of
reaction rate, estimation of the bioreactor performance at a particular operating condition and
certain design is possible. Therefore, for modelling activated sludge reaction rate many specific
software such as BioWin, STOAT, GPS-X and WEST were developed [15].
Bioreactor performance (e.g., the rate of biomass production and effluent substrate
concentration) and design parameters (volume) are estimated by setting up mass balance
equations with using microbial kinetics. Microbial kinetics is largely focus on the microbial
growth rate and substrate utilization [15].

Sewage treatment (or domestic wastewater treatment, municipal wastewater treatment) is a type of wastewater
treatment which aims to remove contaminants from sewage to produce an effluent that is suitable to discharge to the surrounding
environment or an intended reuse application, thereby preventing water pollution from raw sewage discharges.[2] Sewage
contains wastewater from households and businesses and possibly pre-treated industrial wastewater. There are a high number of
sewage treatment processes to choose from. These can range from decentralized systems (including on-site treatment systems) to
large centralized systems involving a network of pipes and pump stations (called sewerage) which convey the sewage to a
treatment plant. For cities that have a combined sewer, the sewers will also carry urban runoff (stormwater) to the sewage treatment
plant. Sewage treatment often involves two main stages, called primary and secondary treatment, while advanced treatment also
incorporates a tertiary treatment stage with polishing processes and nutrient removal. Secondary treatment can reduce organic
matter (measured as biological oxygen demand) from sewage, using aerobic or anaerobic biological processes. A so-called
quarternary treatment step (sometimes referred to as advanced treatment) can also be added for the removal of organic
micropollutants, such as pharmaceuticals. This has been implemented in full-scale for example in Sweden.[3]

A large number of sewage treatment technologies have been developed, mostly using biological treatment processes. Design
engineers and decision makers need to take into account technical and economical criteria of each alternative when choosing a
suitable technology.[4]: 215 Often, the main criteria for selection are: desired effluent quality, expected construction and operating costs,
availability of land, energy requirements and sustainability aspects. In developing countries and in rural areas with low population
densities, sewage is often treated by various on-site sanitation systems and not conveyed in sewers. These systems include septic
tanks connected to drain fields, on-site sewage systems (OSS), vermifilter systems and many more. On the other hand, advanced
and relatively expensive sewage treatment plants may include tertiary treatment with disinfection and possibly even a fourth
treatment stage to remove micropollutants.[3]

At the global level, an estimated 52% of sewage is treated.[5] However, sewage treatment rates are highly unequal for different
countries around the world. For example, while high-income countries treat approximately 74% of their sewage, developing
countries treat an average of just 4.2%.[5]

The treatment of sewage is part of the field of sanitation. Sanitation also includes the management of human waste and solid
waste as well as stormwater (drainage) management.[6] The term sewage treatment plant is often used interchangeably with the
term wastewater treatment plant.[4][page needed][7]

Terminology[edit]

Activated sludge sewage treatment plant in Massachusetts, US

The term sewage treatment plant (STP) (or sewage treatment works) is nowadays often replaced with the term wastewater
treatment plant (WWTP).[7][8] Strictly speaking, the latter is a broader term that can also refer to industrial wastewater treatment.

The terms water recycling center or water reclamation plants are also in use as synonyms.
Purposes and overview[edit]
The overall aim of treating sewage is to produce an effluent that can be discharged to the environment while causing as little water
pollution as possible, or to produce an effluent that can be reused in a useful manner.[9] This is achieved by removing contaminants
from the sewage. It is a form of waste management.

With regards to biological treatment of sewage, the treatment objectives can include various degrees of the following: to transform or
remove organic matter, nutrients (nitrogen and phosphorus), pathogenic organisms, and specific trace organic constituents
(micropollutants).[7]: 548

Some types of sewage treatment produce sewage sludge which can be treated before safe disposal or reuse. Under certain
circumstances, the treated sewage sludge might be termed biosolids and can be used as a fertilizer.

The process that raw sewage goes through before being

released back into surface water.

Sewage characteristics[edit]
This section is an excerpt from Sewage § Concentrations and loads.[edit]

Typical values for physical–chemical characteristics of raw sewage in developing countries have been published as follows: 180
g/person/d for total solids (or 1100 mg/L when expressed as a concentration), 50 g/person/d for BOD (300 mg/L), 100 g/person/d for
COD (600 mg/L), 8 g/person/d for total nitrogen (45 mg/L), 4.5 g/person/d for ammonia-N (25 mg/L) and 1.0 g/person/d for total
phosphorus (7 mg/L).[10]: 57 The typical ranges for these values are: 120–220 g/person/d for total solids (or 700–1350 mg/L when
expressed as a concentration), 40–60 g/person/d for BOD (250–400 mg/L), 80–120 g/person/d for COD (450–800 mg/L), 6–10
g/person/d for total nitrogen (35–60 mg/L), 3.5–6 g/person/d for ammonia-N (20–35 mg/L) and 0.7–2.5 g/person/d for total
phosphorus (4–15 mg/L).[10]: 57

For high income countries, the "per person organic matter load" has been found to be approximately 60 gram of BOD per person
per day.[11] This is called the population equivalent (PE) and is also used as a comparison parameter to express the strength
of industrial wastewater compared to sewage.

Collection[edit]
This section is an excerpt from Sewerage.[edit]

Sewerage (or sewage system) is the infrastructure that conveys sewage or surface runoff (stormwater, meltwater, rainwater) using
sewers. It encompasses components such as receiving drains, manholes, pumping stations, storm overflows, and screening
chambers of the combined sewer or sanitary sewer. Sewerage ends at the entry to a sewage treatment plant or at the point of
discharge into the environment. It is the system of pipes, chambers, manholes, etc. that conveys the sewage or storm water.

In many cities, sewage (or municipal wastewater) is carried together with stormwater, in a combined sewer system, to a sewage
treatment plant. In some urban areas, sewage is carried separately in sanitary sewers and runoff from streets is carried in storm
drains. Access to these systems, for maintenance purposes, is typically through a manhole. During high precipitation periods a
sewer system may experience a combined sewer overflow event or a sanitary sewer overflow event, which forces untreated sewage
to flow directly to receiving waters. This can pose a serious threat to public health and the surrounding environment.

Types of treatment processes[edit]

Sewage can be treated close to where the sewage is created, which may be called a decentralized system or even an on-site
system (on-site sewage facility, septic tanks, etc.). Alternatively, sewage can be collected and transported by a network of pipes and
pump stations to a municipal treatment plant. This is called a centralized system (see also sewerage and pipes and infrastructure).

A large number of sewage treatment technologies have been developed, mostly using biological treatment processes (see list of
wastewater treatment technologies). Very broadly, they can be grouped into high tech (high cost) versus low tech (low cost) options,
although some technologies might fall into either category. Other grouping classifications are intensive or mechanized systems
(more compact, and frequently employing high tech options) versus extensive or natural or nature-based systems (usually using
natural treatment processes and occupying larger areas) systems. This classification may be sometimes oversimplified, because a
treatment plant may involve a combination of processes, and the interpretation of the concepts of high tech and low tech, intensive
and extensive, mechanized and natural processes may vary from place to place.

Low tech, extensive or nature-based processes[edit]

Constructed wetland (vertical flow) at Center for Research and

Training in Sanitation, Belo Horizonte, Brazil Trickling

filter sewage treatment plant at Onça Treatment Plant, Belo Horizonte, Brazil
Examples for more low-tech, often less expensive sewage treatment systems are shown below. They often use little or no energy.
Some of these systems do not provide a high level of treatment, or only treat part of the sewage (for example only the toilet
wastewater), or they only provide pre-treatment, like septic tanks. On the other hand, some systems are capable of providing a good
performance, satisfactory for several applications. Many of these systems are based on natural treatment processes, requiring large
areas, while others are more compact. In most cases, they are used in rural areas or in small to medium-sized communities.

Rural Kansas lagoon on private property

For example, waste stabilization ponds are a low cost treatment option with practically no energy requirements but they require a lot
of land.[4]: 236 Due to their technical simplicity, most of the savings (compared with high tech systems) are in terms of operation and
maintenance costs.[4]: 220–243

 Anaerobic digester types and anaerobic digestion, for example:

 o Upflow anaerobic sludge blanket reactor
o Septic tank
o Imhoff tank
 Constructed wetland (see also biofilters)
 Decentralized wastewater system
 Nature-based solutions
 On-site sewage facility
 Sand filter
 Vermifilter
 Waste stabilization pond with sub-types:[4]: 189
o e.g. Facultative ponds, high rate ponds, maturation ponds
Examples for systems that can provide full or partial treatment for toilet wastewater only:

 Composting toilet (see also dry toilets in general)

 Urine-diverting dry toilet
 Vermifilter toilet

High tech, intensive or mechanized processes[edit]

Aeration tank of activated sludge sewage treatment plant (fine-

bubble diffusers) near Adelaide, Australia
Examples for more high-tech, intensive or mechanized, often relatively expensive sewage treatment systems are listed below. Some
of them are energy intensive as well. Many of them provide a very high level of treatment. For example, broadly speaking,
the activated sludge process achieves a high effluent quality but is relatively expensive and energy intensive.[4]: 239

 Activated sludge systems

 Aerobic treatment system
 Enhanced biological phosphorus removal
 Expanded granular sludge bed digestion
 Filtration
 Membrane bioreactor
 Moving bed biofilm reactor
 Rotating biological contactor
 Trickling filter
 Ultraviolet disinfection

Disposal or treatment options[edit]

There are other process options which may be classified as disposal options, although they can also be understood as basic
treatment options. These include: Application of sludge, irrigation, soak pit, leach field, fish pond, floating plant pond, water
disposal/groundwater recharge, surface disposal and storage.[12]: 138

The application of sewage to land is both: a type of treatment and a type of final disposal.[4]: 189 It leads to groundwater recharge
and/or to evapotranspiration. Land application include slow-rate systems, rapid infiltration, subsurface infiltration, overland flow. It is
done by flooding, furrows, sprinkler and dripping. It is a treatment/disposal system that requires a large amount of land per person.
Design aspects[edit]

Upflow anaerobic sludge blanket (UASB) reactor in Brazil

(picture from a small-sized treatment plant), Center for Research and Training in
Sanitation, Belo Horizonte, Brazil
Population equivalent[edit]
The per person organic matter load is a parameter used in the design of sewage treatment plants. This concept is known
as population equivalent (PE). The base value used for PE can vary from one country to another. Commonly used definitions used
worldwide are: 1 PE equates to 60 gram of BOD per person per day, and it also equals 200 liters of sewage per day.[13] This concept
is also used as a comparison parameter to express the strength of industrial wastewater compared to sewage.

Process selection[edit]
When choosing a suitable sewage treatment process, decision makers need to take into account technical and economical criteria.[4]:
215
Therefore, each analysis is site-specific. A life cycle assessment (LCA) can be used, and criteria or weightings are attributed to
the various aspects. This makes the final decision subjective to some extent.[4]: 216 A range of publications exist to help with
technology selection.[4]: 221 [12][14][15]

In industrialized countries, the most important parameters in process selection are typically efficiency, reliability, and space
requirements. In developing countries, they might be different and the focus might be more on construction and operating costs as
well as process simplicity.[4]: 218

Choosing the most suitable treatment process is complicated and requires expert inputs, often in the form of feasibility studies. This
is because the main important factors to be considered when evaluating and selecting sewage treatment processes are numerous.
They include: process applicability, applicable flow, acceptable flow variation, influent characteristics, inhibiting or refractory
compounds, climatic aspects, process kinetics and reactor hydraulics, performance, treatment residuals, sludge processing,
environmental constraints, requirements for chemical products, energy and other resources; requirements for personnel, operating
and maintenance; ancillary processes, reliability, complexity, compatibility, area availability.[4]: 219

With regards to environmental impacts of sewage treatment plants the following aspects are included in the selection process:
Odors, vector attraction, sludge transportation, sanitary risks, air contamination, soil and subsoil contamination, surface water
pollution or groundwater contamination, devaluation of nearby areas, inconvenience to the nearby population.[4]: 220

Odor control[edit]
Odors emitted by sewage treatment are typically an indication of an anaerobic or septic condition.[16] Early stages of processing will
tend to produce foul-smelling gases, with hydrogen sulfide being most common in generating complaints. Large process plants in
urban areas will often treat the odors with carbon reactors, a contact media with bio-slimes, small doses of chlorine, or circulating
fluids to biologically capture and metabolize the noxious gases.[17] Other methods of odor control exist, including addition of iron
salts, hydrogen peroxide, calcium nitrate, etc. to manage hydrogen sulfide levels.[18]
Energy requirements[edit]
The energy requirements vary with type of treatment process as well as sewage strength. For example, constructed wetlands and
stabilization ponds have low energy requirements.[19] In comparison, the activated sludge process has a high energy consumption
because it includes an aeration step. Some sewage treatment plants produce biogas from their sewage sludge treatment process by
using a process called anaerobic digestion. This process can produce enough energy to meet most of the energy needs of the
sewage treatment plant itself.[7]: 1505

For activated sludge treatment plants in the United States, around 30 percent of the annual operating costs is usually required for
energy.[7]: 1703 Most of this electricity is used for aeration, pumping systems and equipment for the dewatering and drying of sewage
sludge. Advanced sewage treatment plants, e.g. for nutrient removal, require more energy than plants that only achieve primary or
secondary treatment.[7]: 1704

Small rural plants using trickling filters may operate with no net energy requirements, the whole process being driven by gravitational
flow, including tipping bucket flow distribution and the desludging of settlement tanks to drying beds. This is usually only practical in
hilly terrain and in areas where the treatment plant is relatively remote from housing because of the difficulty in managing odors.[20][21]

Co-treatment of industrial effluent[edit]

In highly regulated developed countries, industrial wastewater usually receives at least pretreatment if not full treatment at the
factories themselves to reduce the pollutant load, before discharge to the sewer. The pretreatment has the following two main aims:
Firstly, to prevent toxic or inhibitory compounds entering the biological stage of the sewage treatment plant and reduce its efficiency.
And secondly to avoid toxic compounds from accumulating in the produced sewage sludge which would reduce its beneficial
reuse options. Some industrial wastewater may contain pollutants which cannot be removed by sewage treatment plants. Also,
variable flow of industrial waste associated with production cycles may upset the population dynamics of biological treatment units.
[citation needed]

Design aspects of secondary treatment processes[edit]

Main article: Secondary treatment § Design considerations

A poorly maintained anaerobic treatment pond in Kariba,

Zimbabwe (sludge needs to be removed)
Non-sewered areas[edit]
Urban residents in many parts of the world rely on on-site sanitation systems without sewers, such as septic tanks and pit latrines,
and fecal sludge management in these cities is an enormous challenge.[22]

For sewage treatment the use of septic tanks and other on-site sewage facilities (OSSF) is widespread in some rural areas, for
example serving up to 20 percent of the homes in the U.S.[23]

Available process steps[edit]

Sewage treatment often involves two main stages, called primary and secondary treatment, while advanced treatment also
incorporates a tertiary treatment stage with polishing processes.[13] Different types of sewage treatment may utilize some or all of the
process steps listed below.

Preliminary treatment[edit]
Preliminary treatment (sometimes called pretreatment) removes coarse materials that can be easily collected from the raw sewage
before they damage or clog the pumps and sewage lines of primary treatment clarifiers.
Screening[edit]

Preliminary treatment arrangement at small and medium-sized

sewage treatment plants: Manually-cleaned screens and grit chamber (Jales Treatment Plant, São
Paulo, Brazil)
The influent in sewage water passes through a bar screen to remove all large objects like cans, rags, sticks, plastic packets, etc.
carried in the sewage stream.[24] This is most commonly done with an automated mechanically raked bar screen in modern plants
serving large populations, while in smaller or less modern plants, a manually cleaned screen may be used. The raking action of a
mechanical bar screen is typically paced according to the accumulation on the bar screens and/or flow rate. The solids are collected
and later disposed in a landfill, or incinerated. Bar screens or mesh screens of varying sizes may be used to optimize solids
removal. If gross solids are not removed, they become entrained in pipes and moving parts of the treatment plant, and can cause
substantial damage and inefficiency in the process.[25]: 9

Grit removal[edit]

Preliminary treatment: Horizontal flow grit chambers at a

sewage treatment plant in Juiz de Fora, Minas Gerais, Brazil
Grit consists of sand, gravel, rocks, and other heavy materials. Preliminary treatment may include a sand or grit removal channel or
chamber, where the velocity of the incoming sewage is reduced to allow the settlement of grit. Grit removal is necessary to (1)
reduce formation of deposits in primary sedimentation tanks, aeration tanks, anaerobic digesters, pipes, channels, etc. (2) reduce
the frequency of tank cleaning caused by excessive accumulation of grit; and (3) protect moving mechanical equipment from
abrasion and accompanying abnormal wear. The removal of grit is essential for equipment with closely machined metal surfaces
such as comminutors, fine screens, centrifuges, heat exchangers, and high pressure diaphragm pumps.
Grit chambers come in three types: horizontal grit chambers, aerated grit chambers, and vortex grit chambers. Vortex grit chambers
include mechanically induced vortex, hydraulically induced vortex, and multi-tray vortex separators. Given that traditionally, grit
removal systems have been designed to remove clean inorganic particles that are greater than 0.210 millimetres (0.0083 in), most
of the finer grit passes through the grit removal flows under normal conditions. During periods of high flow deposited grit is
resuspended and the quantity of grit reaching the treatment plant increases substantially.[7]

Flow equalization[edit]

Equalization basins can be used to achieve flow equalization. This is especially useful for combined sewer systems which produce
peak dry-weather flows or peak wet-weather flows that are much higher than the average flows.[7]: 334 Such basins can improve the
performance of the biological treatment processes and the secondary clarifiers.[7]: 334

Disadvantages include the basins' capital cost and space requirements. Basins can also provide a place to temporarily hold, dilute
and distribute batch discharges of toxic or high-strength wastewater which might otherwise inhibit biological secondary treatment
(such was wastewater from portable toilets or fecal sludge that is brought to the sewage treatment plant in vacuum trucks). Flow
equalization basins require variable discharge control, typically include provisions for bypass and cleaning, and may also include
aerators and odor control.[26]

Fat and grease removal[edit]

In some larger plants, fat and grease are removed by passing the sewage through a small tank where skimmers collect the fat
floating on the surface. Air blowers in the base of the tank may also be used to help recover the fat as a froth. Many plants,
however, use primary clarifiers with mechanical surface skimmers for fat and grease removal.

Primary treatment[edit]

Rectangular primary settling tanks at a sewage treatment plant

in Oregon, US
Primary treatment is the "removal of a portion of the suspended solids and organic matter from the sewage".[7]: 11 It consists of
allowing sewage to pass slowly through a basin where heavy solids can settle to the bottom while oil, grease and lighter solids float
to the surface and are skimmed off. These basins are called primary sedimentation tanks or primary clarifiers and typically have a
hydraulic retention time (HRT) of 1.5 to 2.5 hours.[7]: 398 The settled and floating materials are removed and the remaining liquid may
be discharged or subjected to secondary treatment. Primary settling tanks are usually equipped with mechanically driven scrapers
that continually drive the collected sludge towards a hopper in the base of the tank where it is pumped to sludge treatment facilities.
[25]: 9–11

Sewage treatment plants that are connected to a combined sewer system sometimes have a bypass arrangement after the primary
treatment unit. This means that during very heavy rainfall events, the secondary and tertiary treatment systems can be bypassed to
protect them from hydraulic overloading, and the mixture of sewage and storm-water receives primary treatment only.[27]

Primary sedimentation tanks remove about 50–70% of the suspended solids, and 25–40% of the biological oxygen demand (BOD).
[7]: 396

Secondary treatment[edit]
Main article: Secondary treatment
 Simplified process flow diagram for a typical large-scale
treatment plant using the activated sludge process
The main processes involved in secondary sewage treatment are designed to remove as much of the solid material as possible.
[13]
They use biological processes to digest and remove the remaining soluble material, especially the organic fraction. This can be
done with either suspended-growth or biofilm processes. The microorganisms that feed on the organic matter present in the sewage
grow and multiply, constituting the biological solids, or biomass. These grow and group together in the form of flocs or biofilms and,
in some specific processes, as granules. The biological floc or biofilm and remaining fine solids form a sludge which can be settled
and separated. After separation, a liquid remains that is almost free of solids, and with a greatly reduced concentration of pollutants.
[13]

Secondary treatment can reduce organic matter (measured as biological oxygen demand) from sewage, using aerobic or anaerobic
processes. The organisms involved in these processes are sensitive to the presence of toxic materials, although these are not
expected to be present at high concentrations in typical municipal sewage.

Tertiary treatment[edit]

Overall setup for a micro filtration system

Advanced sewage treatment generally involves three main stages, called primary, secondary and tertiary treatment but may also
include intermediate stages and final polishing processes. The purpose of tertiary treatment (also called advanced treatment) is to
provide a final treatment stage to further improve the effluent quality before it is discharged to the receiving water body or reused.
More than one tertiary treatment process may be used at any treatment plant. If disinfection is practiced, it is always the final
process. It is also called effluent polishing. Tertiary treatment may include biological nutrient removal (alternatively, this can be
classified as secondary treatment), disinfection and removal of micropollutants, such as environmental persistent pharmaceutical
pollutants.

Tertiary treatment is sometimes defined as anything more than primary and secondary treatment in order to allow discharge into a
highly sensitive or fragile ecosystem such as estuaries, low-flow rivers or coral reefs.[28] Treated water is sometimes disinfected
chemically or physically (for example, by lagoons and microfiltration) prior to discharge into a stream, river, bay, lagoon or wetland,
or it can be used for the irrigation of a golf course, greenway or park. If it is sufficiently clean, it can also be used for groundwater
recharge or agricultural purposes.

Sand filtration removes much of the residual suspended matter.[25]: 22–23 Filtration over activated carbon, also called carbon
adsorption, removes residual toxins.[25]: 19 Micro filtration or synthetic membranes are used in membrane bioreactors and can also
remove pathogens.[7]: 854

Settlement and further biological improvement of treated sewage may be achieved through storage in large human-made ponds or
lagoons. These lagoons are highly aerobic, and colonization by native macrophytes, especially reeds, is often encouraged.

Disinfection[edit]
Disinfection of treated sewage aims to kill pathogens (disease-causing microorganisms) prior to disposal. It is increasingly effective
after more elements of the foregoing treatment sequence have been completed.[29]: 359 The purpose of disinfection in the treatment of
sewage is to substantially reduce the number of pathogens in the water to be discharged back into the environment or to be reused.
The effectiveness of disinfection depends on the quality of the water being treated (e.g. turbidity, pH, etc.), the type of disinfection
being used, the disinfectant dosage (concentration and time), and other environmental variables. Water with high turbidity will be
treated less successfully, since solid matter can shield organisms, especially from ultraviolet light or if contact times are low.
Generally, short contact times, low doses and high flows all militate against effective disinfection. Common methods of disinfection
include ozone, chlorine, ultraviolet light, or sodium hypochlorite.[25]: 16 Monochloramine, which is used for drinking water, is not used in
the treatment of sewage because of its persistence.

Chlorination remains the most common form of treated sewage disinfection in many countries due to its low cost and long-term
history of effectiveness. One disadvantage is that chlorination of residual organic material can generate chlorinated-organic
compounds that may be carcinogenic or harmful to the environment. Residual chlorine or chloramines may also be capable of
chlorinating organic material in the natural aquatic environment. Further, because residual chlorine is toxic to aquatic species, the
treated effluent must also be chemically dechlorinated, adding to the complexity and cost of treatment.

Ultraviolet (UV) light can be used instead of chlorine, iodine, or other chemicals. Because no chemicals are used, the treated water
has no adverse effect on organisms that later consume it, as may be the case with other methods. UV radiation causes damage to
the genetic structure of bacteria, viruses, and other pathogens, making them incapable of reproduction. The key disadvantages of
UV disinfection are the need for frequent lamp maintenance and replacement and the need for a highly treated effluent to ensure
that the target microorganisms are not shielded from the UV radiation (i.e., any solids present in the treated effluent may protect
microorganisms from the UV light). In many countries, UV light is becoming the most common means of disinfection because of the
concerns about the impacts of chlorine in chlorinating residual organics in the treated sewage and in chlorinating organics in the
receiving water.

As with UV treatment, heat sterilization also does not add chemicals to the water being treated. However, unlike UV, heat can
penetrate liquids that are not transparent. Heat disinfection can also penetrate solid materials within wastewater, sterilizing their
contents. Thermal effluent decontamination systems provide low resource, low maintenance effluent decontamination once
installed.

Ozone (O3) is generated by passing oxygen (O2) through a high voltage potential resulting in a third oxygen atom becoming attached
and forming O3. Ozone is very unstable and reactive and oxidizes most organic material it comes in contact with, thereby destroying
many pathogenic microorganisms. Ozone is considered to be safer than chlorine because, unlike chlorine which has to be stored on
site (highly poisonous in the event of an accidental release), ozone is generated on-site as needed from the oxygen in the ambient
air. Ozonation also produces fewer disinfection by-products than chlorination. A disadvantage of ozone disinfection is the high cost
of the ozone generation equipment and the requirements for special operators. Ozone sewage treatment requires the use of
an ozone generator, which decontaminates the water as ozone bubbles percolate through the tank.

Membranes can also be effective disinfectants, because they act as barriers, avoiding the passage of the microorganisms. As a
result, the final effluent may be devoid of pathogenic organisms, depending on the type of membrane used. This principle is applied
in membrane bioreactors.

Biological nutrient removal[edit]

Nitrification process tank at an activated sludge plant in the

United States
Sewage may contain high levels of the nutrients nitrogen and phosphorus. Typical values for nutrient loads per person and nutrient
concentrations in raw sewage in developing countries have been published as follows: 8 g/person/d for total nitrogen (45 mg/L), 4.5
g/person/d for ammonia-N (25 mg/L) and 1.0 g/person/d for total phosphorus (7 mg/L).[4]: 57 The typical ranges for these values are: 6-
10 g/person/d for total nitrogen (35–60 mg/L), 3.5-6 g/person/d for ammonia-N (20–35 mg/L) and 0.7-2.5 g/person/d for total
phosphorus (4–15 mg/L).[4]: 57

Excessive release to the environment can lead to nutrient pollution, which can manifest itself in eutrophication. This process can
lead to algal blooms, a rapid growth, and later decay, in the population of algae. In addition to causing deoxygenation, some algal
species produce toxins that contaminate drinking water supplies.

Ammonia nitrogen, in the form of free ammonia (NH3) is toxic to fish. Ammonia nitrogen, when converted to nitrite and further to
nitrate in a water body, in the process of nitrification, is associated with the consumption of dissolved oxygen. Nitrite and nitrate may
also have public health significance if concentrations are high in drinking water, because of a disease called metahemoglobinemia.[4]:
42
Phosphorus removal is important as phosphorus is a limiting nutrient for algae growth in many fresh water systems. Therefore, an
excess of phosphorus can lead to eutrophication. It is also particularly important for water reuse systems where high phosphorus
concentrations may lead to fouling of downstream equipment such as reverse osmosis.

A range of treatment processes are available to remove nitrogen and phosphorus. Biological nutrient removal (BNR) is regarded by
some as a type of secondary treatment process,[7] and by others as a tertiary (or advanced) treatment process.

Nitrogen removal[edit]

Constructed wetlands (vertical flow) for sewage treatment

near Shanghai, China
Nitrogen is removed through the biological oxidation of nitrogen from ammonia to nitrate (nitrification), followed by denitrification, the
reduction of nitrate to nitrogen gas. Nitrogen gas is released to the atmosphere and thus removed from the water.

Nitrification itself is a two-step aerobic process, each step facilitated by a different type of bacteria. The oxidation of ammonia (NH4+)
to nitrite (NO2−) is most often facilitated by bacteria such as Nitrosomonas spp. (nitroso refers to the formation of a nitroso functional
group). Nitrite oxidation to nitrate (NO3−), though traditionally believed to be facilitated by Nitrobacter spp. (nitro referring the
formation of a nitro functional group), is now known to be facilitated in the environment predominantly by Nitrospira spp.

Denitrification requires anoxic conditions to encourage the appropriate biological communities to form. Anoxic conditions refers to a
situation where oxygen is absent but nitrate is present. Denitrification is facilitated by a wide diversity of bacteria. The activated
sludge process, sand filters, waste stabilization ponds, constructed wetlands and other processes can all be used to reduce
nitrogen.[25]: 17–18 Since denitrification is the reduction of nitrate to dinitrogen (molecular nitrogen) gas, an electron donor is needed.
This can be, depending on the wastewater, organic matter (from the sewage itself), sulfide, or an added donor like methanol. The
sludge in the anoxic tanks (denitrification tanks) must be mixed well (mixture of recirculated mixed liquor, return activated sludge,
and raw influent) e.g. by using submersible mixers in order to achieve the desired denitrification.

Over time, different treatment configurations for activated sludge processes have evolved to achieve high levels of nitrogen removal.
An initial scheme was called the Ludzack–Ettinger Process. It could not achieve a high level of denitrification.[7]: 616 The Modified
Ludzak–Ettinger Process (MLE) came later and was an improvement on the original concept. It recycles mixed liquor from the
discharge end of the aeration tank to the head of the anoxic tank. This provides nitrate for the facultative bacteria.[7]: 616

There are other process configurations, such as variations of the Bardenpho process.[30]: 160 They might differ in the placement of
anoxic tanks, e.g. before and after the aeration tanks.

Phosphorus removal[edit]

Studies of United States sewage in the late 1960s estimated mean per capita contributions of 500 grams (18 oz) in urine and feces,
1,000 grams (35 oz) in synthetic detergents, and lesser variable amounts used as corrosion and scale control chemicals in water
supplies.[31] Source control via alternative detergent formulations has subsequently reduced the largest contribution, but naturally the
phosphorus content of urine and feces remained unchanged.

Phosphorus can be removed biologically in a process called enhanced biological phosphorus removal. In this process, specific
bacteria, called polyphosphate-accumulating organisms (PAOs), are selectively enriched and accumulate large quantities of
phosphorus within their cells (up to 20 percent of their mass).[30]: 148–155

Phosphorus removal can also be achieved by chemical precipitation, usually with salts of iron (e.g. ferric chloride)
or aluminum (e.g. alum), or lime.[25]: 18 This may lead to a higher sludge production as hydroxides precipitate and the added chemicals
can be expensive. Chemical phosphorus removal requires significantly smaller equipment footprint than biological removal, is easier
to operate and is often more reliable than biological phosphorus removal. Another method for phosphorus removal is to use
granular laterite or zeolite.[32][33]
Some systems use both biological phosphorus removal and chemical phosphorus removal. The chemical phosphorus removal in
those systems may be used as a backup system, for use when the biological phosphorus removal is not removing enough
phosphorus, or may be used continuously. In either case, using both biological and chemical phosphorus removal has the
advantage of not increasing sludge production as much as chemical phosphorus removal on its own, with the disadvantage of the
increased initial cost associated with installing two different systems.

Once removed, phosphorus, in the form of a phosphate-rich sewage sludge, may be sent to landfill or used as fertilizer in admixture
with other digested sewage sludges. In the latter case, the treated sewage sludge is also sometimes referred to as biosolids. 22% of
the world's phosphorus needs could be satisfied by recycling residential wastewater.[34][35]

Fourth treatment stage[edit]

Further information: Environmental impact of pharmaceuticals and personal care products
Micropollutants such as pharmaceuticals, ingredients of household chemicals, chemicals used in small businesses or
industries, environmental persistent pharmaceutical pollutants (EPPP) or pesticides may not be eliminated in the commonly used
sewage treatment processes (primary, secondary and tertiary treatment) and therefore lead to water pollution.[36] Although
concentrations of those substances and their decomposition products are quite low, there is still a chance of harming aquatic
organisms. For pharmaceuticals, the following substances have been identified as toxicologically relevant: substances
with endocrine disrupting effects, genotoxic substances and substances that enhance the development of bacterial resistances.
[37]
They mainly belong to the group of EPPP.

Techniques for elimination of micropollutants via a fourth treatment stage during sewage treatment are implemented in Germany,
Switzerland, Sweden[3] and the Netherlands and tests are ongoing in several other countries.[38] Such process steps mainly consist
of activated carbon filters that adsorb the micropollutants. The combination of advanced oxidation with ozone followed by granular
activated carbon (GAC) has been suggested as a cost-effective treatment combination for pharmaceutical residues. For a full
reduction of microplasts the combination of ultrafiltration followed by GAC has been suggested. Also the use of enzymes such
as laccase secreted by fungi is under investigation.[39][40] Microbial biofuel cells are investigated for their property to treat organic
matter in sewage.[41]

To reduce pharmaceuticals in water bodies, source control measures are also under investigation, such as innovations in drug
development or more responsible handling of drugs.[37][42] In the US, the National Take Back Initiative is a voluntary program with the
general public, encouraging people to return excess or expired drugs, and avoid flushing them to the sewage system.[43]

Sludge treatment and disposal[edit]

View of a belt filter press at the Blue Plains Advanced

Wastewater Treatment Plant, Washington, D.C. Mechanical

dewatering of sewage sludge with a centrifuge at a large sewage treatment plant (Arrudas
Treatment Plant, Belo Horizonte, Brazil)
This section is an excerpt from Sewage sludge treatment.[edit]

Sewage sludge treatment describes the processes used to manage and dispose of sewage sludge produced during sewage
treatment. Sludge treatment is focused on reducing sludge weight and volume to reduce transportation and disposal costs, and on
reducing potential health risks of disposal options. Water removal is the primary means of weight and volume reduction,
while pathogen destruction is frequently accomplished through heating during thermophilic digestion, composting, or incineration.
The choice of a sludge treatment method depends on the volume of sludge generated, and comparison of treatment costs required
for available disposal options. Air-drying and composting may be attractive to rural communities, while limited land availability may
make aerobic digestion and mechanical dewatering preferable for cities, and economies of scale may encourage energy
recovery alternatives in metropolitan areas.

Sludge is mostly water with some amounts of solid material removed from liquid sewage. Primary sludge includes settleable
solids removed during primary treatment in primary clarifiers. Secondary sludge is sludge separated in secondary clarifiers that are
used in secondary treatment bioreactors or processes using inorganic oxidizing agents. In intensive sewage treatment processes,
the sludge produced needs to be removed from the liquid line on a continuous basis because the volumes of the tanks in the liquid
line have insufficient volume to store sludge.[44] This is done in order to keep the treatment processes compact and in balance
(production of sludge approximately equal to the removal of sludge). The sludge removed from the liquid line goes to the sludge
treatment line. Aerobic processes (such as the activated sludge process) tend to produce more sludge compared with anaerobic
processes. On the other hand, in extensive (natural) treatment processes, such as ponds and constructed wetlands, the produced
sludge remains accumulated in the treatment units (liquid line) and is only removed after several years of operation.[45]

Sludge treatment options depend on the amount of solids generated and other site-specific conditions. Composting is most often
applied to small-scale plants with aerobic digestion for mid-sized operations, and anaerobic digestion for the larger-scale operations.
The sludge is sometimes passed through a so-called pre-thickener which de-waters the sludge. Types of pre-thickeners include
centrifugal sludge thickeners,[46] rotary drum sludge thickeners and belt filter presses.[47] Dewatered sludge may be incinerated or
transported offsite for disposal in a landfill or use as an agricultural soil amendment.[48]

Environmental impacts[edit]
Sewage treatment plants can have significant effects on the biotic status of receiving waters and can cause some water pollution,
especially if the treatment process used is only basic. For example, for sewage treatment plants without nutrient
removal, eutrophication of receiving water bodies can be a problem.

This section is an excerpt from Water pollution.[edit]

Water pollution (or aquatic pollution) is the contamination of water bodies, usually as a result of human activities, so that it negatively
affects its uses.[49]: 6 Water bodies include lakes, rivers, oceans, aquifers, reservoirs and groundwater. Water pollution results
when contaminants mix with these water bodies. Contaminants can come from one of four main sources: sewage discharges,
industrial activities, agricultural activities, and urban runoff including stormwater.[50] Water pollution is either surface water pollution
or groundwater pollution. This form of pollution can lead to many problems, such as the degradation of aquatic ecosystems or
spreading water-borne diseases when people use polluted water for drinking or irrigation.[51] Another problem is that water pollution
reduces the ecosystem services (such as providing drinking water) that the water resource would otherwise provide.

Treated effluent from sewage treatment plant in Děčín, Czech

Republic, is discharged to surface waters.

Reuse[edit]
Further information: Reuse of excreta
 Sludge drying beds for sewage sludge treatment at a small
treatment plant at the Center for Research and Training in Sanitation, Belo Horizonte, Brazil
Irrigation[edit]
See also: Sewage farm
Increasingly, people use treated or even untreated sewage for irrigation to produce crops. Cities provide lucrative markets for fresh
produce, so are attractive to farmers. Because agriculture has to compete for increasingly scarce water resources with industry and
municipal users, there is often no alternative for farmers but to use water polluted with sewage directly to water their crops. There
can be significant health hazards related to using water loaded with pathogens in this way. The World Health
Organization developed guidelines for safe use of wastewater in 2006.[52] They advocate a 'multiple-barrier' approach to wastewater
use, where farmers are encouraged to adopt various risk-reducing behaviors. These include ceasing irrigation a few days before
harvesting to allow pathogens to die off in the sunlight, applying water carefully so it does not contaminate leaves likely to be eaten
raw, cleaning vegetables with disinfectant or allowing fecal sludge used in farming to dry before being used as a human manure.[53]

Circular secondary sedimentation tank at activated

sludge sewage treatment plant at Arrudas Treatment Plant, Belo Horizonte, Brazil
Reclaimed water[edit]
This section is an excerpt from Reclaimed water.[edit]

Water reclamation (also called wastewater reuse, water reuse or water recycling) is the process of converting municipal
wastewater (sewage) or industrial wastewater into water that can be reused for a variety of purposes. Types of reuse include: urban
reuse, agricultural reuse (irrigation), environmental reuse, industrial reuse, planned potable reuse, and de facto wastewater reuse
(unplanned potable reuse). For example, reuse may include irrigation of gardens and agricultural fields or replenishing surface
water and groundwater (i.e., groundwater recharge). Reused water may also be directed toward fulfilling certain needs in residences
(e.g. toilet flushing), businesses, and industry, and could even be treated to reach drinking water standards. The injection of
reclaimed water into the water supply distribution system is known as direct potable reuse. However, drinking reclaimed water is not
a typical practice.[54] Treated municipal wastewater reuse for irrigation is a long-established practice, especially in arid countries.
Reusing wastewater as part of sustainable water management allows water to remain as an alternative water source for human
activities. This can reduce scarcity and alleviate pressures on groundwater and other natural water bodies.[55]
Global situation[edit]

Share of domestic wastewater that is safely treated (in 2018) [56]

Before the 20th century in Europe, sewers usually discharged into a body of water such as a river, lake, or ocean. There was no
treatment, so the breakdown of the human waste was left to the ecosystem. This could lead to satisfactory results if the assimilative
capacity of the ecosystem is sufficient which is nowadays not often the case due to increasing population density.[4]: 78

Today, the situation in urban areas of industrialized countries is usually that sewers route their contents to a sewage treatment plant
rather than directly to a body of water. In many developing countries, however, the bulk of municipal and industrial wastewater is
discharged to rivers and the ocean without any treatment or after preliminary treatment or primary treatment only. Doing so can lead
to water pollution. Few reliable figures exist on the share of the wastewater collected in sewers that is being treated in the world. A
global estimate by UNDP and UN-Habitat in 2010 was that 90% of all wastewater generated is released into the environment
untreated.[57] A more recent study in 2021 estimated that globally, about 52% of sewage is treated.[5] However, sewage treatment
rates are highly unequal for different countries around the world. For example, while high-income countries treat approximately 74%
of their sewage, developing countries treat an average of just 4.2%.[5] As of 2022, without sufficient treatment, more than 80% of all
wastewater generated globally is released into the environment. High-income nations treat, on average, 70% of the wastewater they
produce, according to UN Water.[34][58][59] Only 8% of wastewater produced in low-income nations receives any sort of treatment.[34][60][61]

The Joint Monitoring Programme (JMP) for Water Supply and Sanitation by WHO and UNICEF report in 2021 that 82% of people
with sewer connections are connected to sewage treatment plants providing at least secondary treatment.[62]: 55 However, this value
varies widely between regions. For example, in Europe, North America, Northern Africa and Western Asia, a total of 31 countries
had universal (>99%) wastewater treatment. However, in Albania, Bermuda, North Macedonia and Serbia "less than 50% of
sewered wastewater received secondary or better treatment" and in Algeria, Lebanon and Libya the value was less than 20% of
sewered wastewater that was being treated. The report also found that "globally, 594 million people have sewer connections that
don't receive sufficient treatment. Many more are connected to wastewater treatment plants that do not provide effective treatment
or comply with effluent requirements.".[62]: 55

Global targets[edit]
Sustainable Development Goal 6 has a Target 6.3 which is formulated as follows: "By 2030, improve water quality by reducing
pollution, eliminating dumping and minimizing release of hazardous chemicals and materials, halving the proportion of untreated
wastewater and substantially increasing recycling and safe reuse globally."[56] The corresponding Indicator 6.3.1 is the "proportion of
wastewater safely treated". It is anticipated that wastewater production would rise by 24% by 2030 and by 51% by 2050.[34][63][64]

Data in 2020 showed that there is still too much uncollected household wastewater: Only 66% of all household wastewater flows
were collected at treatment facilities in 2020 (this is determined from data from 128 countries).[8]: 17 Based on data from 42 countries
in 2015, the report stated that "32 per cent of all wastewater flows generated from point sources received at least some treatment".[8]:
17
For sewage that has indeed been collected at centralized sewage treatment plants, about 79% went on to be safely treated in
2020.[

PRRIVIEWED DOCUMENT

Optimizing the efficiency of urban water systems is a growing concern for water utilities worldwide. Wastewater
treatment plants (WWTP) are crucial in maintaining water quality and resource recovery in a world facing
growing challenges such as climate change, water-energy-food nexus and the increase of legal requirements
and users' expectations. Thus, adopting a performance assessment system (PAS) is of the utmost importance to
assess operating conditions and to identify critical aspects of the WWTP which can negatively affect its
effectiveness, efficiency and reliability. This paper presents the global and operational performance assessment
of an urban WWTP and identifies improvement measures. The WWTP presented a good performance in terms
of effectiveness and reliability. Nevertheless, in terms of efficiency, relevant improvement opportunities were
identified, specifically in the sludge treatment phase and in terms of energy management. PAS was proven to
be successful in the identification and prioritization of rehabilitation needs in a systematic way which will
continuously improve the efficiency and effectiveness of the WWTP as well as to support asset management
decisions regarding their upgrade and retrofit.

Highlights
Listen

 A performance assessment system methodology was applied to an urban WWTP.

 The global and operational performance assessment based on indicators and indices is presented.
 The WWTP presented a good overall performance in terms of effectiveness.
 The performance assessment system allowed the identification of efficiency improvement measures.
key performance indicators, operational efficiency and effectiveness, performance indices, wastewater
treatment plants

INTRODUCTION
Listen
Wastewater treatment plants (WWTP) play an important role in maintaining water quality and resource
recovery in a world facing significant challenges such as climate change, water-energy-food nexus and the
increase of legal requirements and user's expectations. As such, maintaining efficient systems is a growing
concern for water utilities. However, WWTP are typically designed to solely meet wastewater quality discharge
permits (effectiveness) and are not traditionally designed to maintain costs at a minimum (efficiency)
(Panepinto et al. 2016 and references therein). Good management of the WWTP is imperative to balance water
pollution prevention and economic sustainability of these facilities. Adopting a preventive management will
allow the achievement of the required levels of service (reliability) whilst optimally utilizing the available
resources (efficiency). For this, a ‘plan-do-check-act’ approach is usually recommended (Cabrera et
al. 2011; Silva et al. 2014b) for which performance assessment is of most importance. Thus, adopting a
methodology for the performance assessment of WWTP is crucial to evaluate its efficiency and effectiveness as
a starting point to identify the strengths and improvement measures in a continuous way.
Several performance assessment methodologies have been proposed and are typically based on life cycle
analysis (LCA) and energy efficiency benchmarking. LCA methodologies mainly evaluate the potential
environmental impact of products and services of WWTP, helping utility managers to maintain a sustainable
water management (Arnell et al. 2017 and references therein). Energy benchmarking aims at providing key
performance indicators (KPI) and methodologies to help increase energy efficiency diagnosis (Yang et
al. 2010; Longo et al. 2016 and references therein; Panepinto et al. 2016; Di Fraia et al. 2018; Longo et
al. 2019; Luo et al. 2019). Other benchmarking tools are based on assessing the effectiveness and economic
efficiency of WWTP by analysing the effluent characteristics to guarantee compliance using KPIs and to identify
potential economic savings (Molinos-Senante et al. 2014; Ebrahimi et al. 2017).
Even though these studies have brought great contributions to the field, typically they assess efficiency on a
yearly basis. In addition, they lack a general overview and assessment of the WWTP in terms of both
effectiveness and efficiency of energy and sludge management and adequacy of operational practice in each of
the WWTP processes. Such an integrated assessment will allow the identification of improvement needs in
terms of technical, economic and environmental performance (Silva et al. 2014a). In addition, the use of KPI is
crucial as it allows understanding of the overall effectiveness and efficiency of a WWTP. However, such analyses
fail at providing crucial information on the complexity of such systems. Individual processes in a WWTP will be
influenced differently by dynamic parameters such as inlet flow, organic matter and nutrients concentrations or
solid content (Di Fraia et al. 2018 and references therein). Thus, although KPI allow for an adequate assessment
of any system, they should be used as a starting point for a more specific and detailed analysis of each element
(Okwori et al. 2020 and references therein) or process.
AGS, owned by Marubeni, is a privately held company responsible for the operation and maintenance of
several water and wastewater treatment facilities and for the management of 13 utilities in Portugal and Brazil
under concession agreements, public–private partnerships and for the service provision of engineering services
to water utilities in Europe, South America and Asia. Aiming at greater efficiency and effectiveness, AGS has
applied a performance assessment system (PAS) (Rosa et al. 2011; Silva et al. 2014a, 2016), under the scope of
the Portuguese benchmarking on Water Quality, Treatment and Energy initiative (iEQTA) (Silva & Rosa 2020), to
evaluate the performance of several WWTP over the past years. Briefly, this methodology takes into account the
overall effectiveness of WWTP to guarantee compliance and to identify areas which can be improved.
Subsequently, a more detailed analysis of each process is performed which allows the pinpointing of specific
improvements in the operation of each process of the WWTP. This paper describes the application of this
methodology to a selected WWTP. Subsequently, the global and operational performance assessment is
presented and improvement measures for this specific WWTP are identified.

METHODS
Performance assessment system
Listen
The PAS methodology was applied to a WWTP in order to obtain an indication of how the plant could be
optimized in terms of efficiency and effectiveness. Briefly, the methodology involves two stages (global and
operational performance assessment). The first consists of evaluating the WWTP's global performance (GP)
using KPI in terms of its removal efficiency and reliability, energy efficiency and sludge management on an
annual basis. The KPI selected to evaluate the chosen WWTP are depicted in Table 1.
Table 1
Key performance indicators for the global performance assessment of the wastewater treatment plant
Goal Code, description and units Formula
Effectiveness and WQ01 – Quality tests carried out (discharge permit regulation) (%)
reliability

WQ02 – Parameters analysed (discharge permit regulation) (%)

WQ03.1 – Compliance of discharged water quality discharge

permit regulation (%)
where m is required parameters analysed
WQ03.3 – Compliance with Portuguese legislation (%)

where p is parameters analysed (no.); q is

no compliance or 1 – compliance)
ER01 – Volumetric efficiency (%)

Energy efficiency ER08 – Net use of energy from external sources (kWh/m ) 3

RU03.1 – Energy consumption (kWh/m ) 3

RU03.2 – Energy consumption (kWh/kg BOD removed)

5

BP18 – Energy production from biogas (kWh/m ) 3

Sludge management BP01.1 – Sludge production (kg/m ) 3

Goal Code, description and units Formula
BP01.2 – Sludge production (kg/kg BOD removed)
5

BP08 – Sludge dry weight (%) Percentage of dry weight in the produced
The second stage evaluates the system in terms of the daily operational performance of treated wastewater
quality, removal efficiency and operational conditions. Each individual operation/process is evaluated using
performance indices (PX) which can vary between 0 and 300 where a PX of 100 corresponds to the minimum
acceptable performance and 300 corresponds to an excellent performance. Values between 0 and 100
correspond to poor performance, values between 100 and 200 correspond to acceptable performance and
values between 200 and 300 correspond to good performance. The PX are obtained by converting state-
variable data, which express the relevant operational performance assessment aspects of the WWTP into
dimensionless performance indices (Silva et al. 2016 and references therein) using performance functions
(Figure 1). The type of performance function used depends on which parameter is being assessed (view
subsections below).
Figure 1

VIEW LARGEDOWNLOAD SLIDE

Performance functions used converting state variables of relevant operational performance assessment
aspects of the WWTP into dimensionless performance indices. (R0 to R300 are reference values) (red – poor
performance, yellow – acceptable performance, green – good performance).
Treated wastewater quality
Listen
The treated wastewater quality PX are determined using a performance function (Figure 1) where R100 is the
parametric value (PV), R200 corresponds to half of the PV, i.e. 0.5 PV, and R300 corresponds to 0.2 PV. R0 is
determined according to the maximum deviation allowed by the legislation for each parameter. For example,
R0 is 2 PV for chemical oxygen demand (COD) and biochemical oxygen demand (BOD 5) and 2.5 PV for total
suspended solids (TSS). PX for COD, BOD5 and TSS are determined using the performance function shown
in Figure 1(a) whereas for pH, the performance function used is shown in Figure 1(b) since it is dependent on a
range of values.
Removal efficiency
Listen
Assessing facilities' performance based on typical removal efficiencies (Er) (Equation (1)) can be misleading since
a given Er can either be insufficient, adequate or excessive to achieve effluent quality requirements, depending
on influent quality.

(1)
where Cout is the effluent concentration and Cin is the influent concentration.
Thus, to obtain a clear picture on how effective removal efficiencies are, PX on removal efficiencies (PX_Er) were
determined according to the methodology presented in Silva et al. (2014b). The reference values and
performance functions depend on the targeted pollutants and the specific operation or process and take into
consideration the influent concentration (Cin) and limit value for effluent concentration, field data of Er vs Cin (to
obtain the Er model curves) and lower limit of Er vs Cin typical curves which are determined based on literature
review, taking into account the type of treatment. Consequently, the PX_Er are obtained by converting a 0–100%
Er into a 0–300 performance indices using the increasing performance function depicted in Figure 1(c). The
performance functions used for PX_Er were obtained for COD, BOD 5 and TSS removal considering a WWTP with
activated sludge followed by secondary clarifiers.
Operational conditions
Listen
The evaluation of operational conditions identifies the key state operational parameters for each
process/treatment step which can be limiting the overall performance. Regarding operational conditions, PX
100 is obtained when the parameter values are within the range suggested in the literature. On the other hand,
PX 200 corresponds to the values which guarantee the commitment between removal and economic efficiency.
PX300 was not considered in most operational parameters as it cannot be based on literature values since
excellent performance is very specific to each WWTP and its operational conditions. Thus, the assessment must
identify case by case if there is potential for improvement or not (Quadros et al. 2010a).
Table 2 summarizes the specific key state operational parameters selected for the main treatment units in the
studied WWTP and the reference values which were considered in the present study. The parameters were
selected based on the specificity of the WWTP process and the available data.
Table 2
State variables/operational parameters selected to evaluate the operational conditions of the WWTP
Treatment unit Operational parameter Performance function type Reference values

R0 R100 R200

Primary clarifier (no addition of coagulant) HRT (d) Figure 1(b) 0.75 1 1.5
Q/A (m /m .d)
3 2
Figure 1(b) 18.75 25 30
Activated sludge HRT_anaerobic (d) Figure 1(b) 0.3 0.5 0.7
HRT_anoxic (d) Figure 1(b) 0.4 0.5 0.6
HRT_aerobic (d) Figure 1(b) 3 3.5 4
R (%)
sludge Figure 1(b) 16 20 25
R (%)
I Figure 1(b) 80 100 120
MLSS (mg/L) Figure 1(b) 1,600 2,000 3,000
F/M (kgBOD /kgMLSS/d)
5 Figure 1(b) 0.13 0.15 0.18
SRT (d) Figure 1(b) 3 4 7
Secondary clarifier Q/A (m /m .d)
3 2
Figure 1(b) 11 15 20
SLR (kgTSS/(m .h)) 2
Figure 1(b) 1.5 2 4
HRT (d) Figure 1(b) 1.4 1.5 1.6
Anaerobic digestion DM (% w/w) Figure 1(c) 1.125 1.5 2.5
Alkalinity (mg CaCO /L) 3 Figure 1(b) 1,500 2,000 3,000
pH Figure 1(b) 6 6.6 7
VFA/Alkalinity (mg VFA/mgCaCO ) Figure 1(b)
3 0.075 0.1 0.2
Gravitational thickener DM (% w/w) Figure 1(c) 3.75 5 n.a.
Q/A (m /m .d)
3 2
Figure 1(b) 11.625 15.5 23.75
SLR (kgDM/(m .h)) 2
Figure 1(b) 67.5 90 120
HRT (d) Figure 1(b) 0.75 1 1.2
Flotation DM (% w/w) Figure 1(c) 2.25 3 4
Dewatering (centrifuge) DM (% w/w) Figure 1(b) 11.25 15 17.5
HRT – Hydraulic retention time; Q/A – Overflow rate; RSludge – Sludge recirculation; RI – Internal recirculation;
MLSS – Mixed liquor suspended solids; F/M – Food to microorganism ratio; SRT – Solids retention time; SLR –
Solids loading rate; DM – Dry matter; VFA – Volatile fatty acids; n.a. non-applicable.

RESULTS AND DISCUSSION

Case study
Listen
As a case study, a WWTP serving a 263,107 population equivalent with a treated flow of 27,922 m 3/day and an
organic loading of 13,984 kgBOD5/day is presented. The wastewater treatment includes primary clarification
(two units), secondary activated sludge process (A2O – anaerobic/anoxic/aerobic) and secondary clarification
(three units). The produced sludge is treated by gravitational thickening and flotation, anaerobic digestion with
biogas production and centrifuge sludge dewatering.
The WWTP performance was evaluated for the period 2015–2018.

Global performance assessment

Listen
The GP of the WWTP can be seen in Figure 2. Although an overall satisfactory performance is observed, the GP
gave an indication that improvement is possible in all areas, especially in terms of energy performance and
sludge management.
Figure 2

VIEW LARGEDOWNLOAD SLIDE

WWTP's global performance assessment (red – poor performance, yellow – acceptable performance, green –
good performance, grey – non-applicable).
The KPI for effectiveness and reliability show that the WWTP consistently complies with the discharge permit
regulation. Note that WQ3.3 was determined using all the operational control analyses and not only those
required by the Portuguese regulation. Nevertheless, this WWTP shows efficiency problems by not consistently
ensuring compliance in terms of BOD5, COD and TSS (view section below). This is partly due to the discharge of
non-permitted industrial wastewater to the WWTP.
Since WWTP are responsible for 1% of electricity consumption in European countries (Simon-Várhelyi et
al. 2020) and 25–40% of operating costs in a WWTP are related to energy consumption (Panepinto et
al. 2016 and references therein), it is crucial to evaluate the energy efficiency and to identify improvement
measures. Regarding energy production, the results obtained for BP18 show that the WWTP could be
producing higher volumes of biogas which could be converted to energy. Improving the parameter and
decreasing energy consumption will have a direct impact on the results obtained for ER08. In terms of energy
performance, RU03.2 presents good performance, showing an adequate energy efficiency in terms of mass
removed. However, the performance indicator for energy consumption (kWh/m 3), RU03.1, presents poor
performance resulting in an average consumption of 0.58 kWh/m 3. The result obtained is similar to that
estimated for energy consumption (0.6 kWh/m3) of a typical WWTP with activated sludge processes and
anaerobic digestion of the sludge (Gude 2015). As such, and based on these results, energy measurement
campaigns were identified as necessary and are currently being performed to understand which
units/equipment are underperforming and can be optimized. These energy campaigns will focus on the main
energy consuming equipment such as aeration of mixed liquor sludge pumping in primary and secondary
settling and sludge dewatering which typically represent 55–70, 15.6 and 7% of total energy consumption,
respectively (Metcalf & Eddy 2003).
The main indicators affecting the sludge management efficiency are the sludge dry weight (BP08) and sludge
production (BP01.1), which indicates possible problems in the sludge treatment units (see section below).
However, an improvement can be observed in the KPI BP01.2, which went from an acceptable performance in
2015–2016 to a good performance in 2017–2018, and KPI BP01.1, which went from a poor performance in
2015–2016 to an acceptable performance in 2017–2018.

Operational performance
Quality of the treated wastewater
Listen
Overall, the PX of BOD5 and COD show that the WWTP can remove these pollutants to levels which are in
compliance with the discharge permit (Figure 3). The lower performance around May 2018 was due to non-
permitted industrial wastewater discharge and an anomalous increase in rainfall which increased the inflow. If
recurrent, such non-permitted industrial discharges should be evaluated in terms of origin and characteristics
to minimize the impact in the treatment effectiveness (Sánchez-Avila et al. 2009).
On the other hand, TSS values showed great variability throughout the years of study (Figure 3). Albeit the
effluent concentration values are compliant with the discharge permit, and this variability may indicate that
there are some instabilities in the operational conditions which must be studied further to identify
improvement measures and minimize the risks.
Figure 3

VIEW LARGEDOWNLOAD SLIDE

Daily performance indices of treated wastewater quality for COD, BOD 5 and TSS.
Removal efficiency
Listen
Figure 4 presents the results obtained for the removal efficiency of COD, BOD 5 and TSS for the studied WWTP.
Despite receiving wastewater with concentrations above the typical range in terms of COD and BOD 5 (Figure 4,
left), the WWTP has been successful at removing these pollutants, keeping removal at an acceptable or good
performance level (Figure 4, right). The only exception was the moment corresponding to the industrial
wastewater discharge as mentioned above. Contrarywise, TSS levels in the influent were mostly within the
typical range but showed lower performance levels, i.e. lower PX_Er. This is coherent with the results obtained
in terms of quality shown above. As such, the operational conditions of the primary and secondary clarifier
were studied and are presented and discussed in the following section.
Figure 4

VIEW LARGEDOWNLOAD SLIDE

Er model curves (for theoretical and real Er), effect of Cin on Er and typical Cin range (left), examples of
performance functions used converting removal efficiencies of COD, BOD 5 and TSS into PX_Er (middle) and
PX_Er obtained for the studied WWTP (right).
Operating conditions
Listen
To understand where the system is underperforming, each operational unit was studied using the data
available in order to calculate the operational PX (Table 2). Since the main problems detected are related with
the TSS (Figure 3) and sludge management (Figure 2), this section focuses mainly on the operational
parameters associated with these issues.
The PX for surface overflow rate and HRT in the primary clarifiers were mostly below 100 (in the red area) (data
not shown). This is mainly the case in the summer months where the inflow is lower. Most likely this occurs
since the inflow is consistently lower than what was expected when designing the WWTP. Since the WWTP has
two primary clarifiers, and given that the results indicate oversizing of the primary clarification phase, it could
be advantageous to disable one of the units.
The secondary clarifiers also present potential for improvement (Figure 5). The performance indices show that
there is no consistency in these operational parameters. The sludge recirculation rate (R Sludge) is crucial to
maintain a sufficient concentration of activated sludge in the aeration tank and the sludge blanket depth in the
secondary clarifier (Metcalf & Eddy 2003). Currently, the secondary clarifier and the activated sludge unit are
being operated at high RSludge. As such, it is suggested to decrease the RSludge to values between 25 and 50%
(Table 2). This will have a direct impact on the SLR and on the HRT. Similarly to the primary clarifiers, the low
performance indices for the overflow rate are likely to be related with the oversizing of the WWTP. As such, it
could be advantageous to work with only one or two units of the secondary clarifiers. By maintaining the best
sludge recirculation rate would assure optimal operation of the secondary clarifiers, and it would be possible to
maintain a stable high-quality effluent whilst reducing operational costs (Conserva et al. 2019).
In terms of sludge management and analysing the dry solids PX, a good performance was observed for all
stages of sludge treatment except for the sludge thickening phase which is presented in Figure 6. This
underperformance could be associated with the low PX obtained for hydraulic and solids loading rates and
hydraulic retention time (HRT) in the gravitational sludge thickening unit (Figure 6). In fact, the low hydraulic
loading rates (Figure 6(b)) and high HRT observed (Figure 6(d)) can lead to floating sludge which may reduce the
thickening efficiency (Metcalf & Eddy 2003). Since sludge management can cost between 20 and 60% of the
total operating costs in wastewater treatment (IWA 2007), careful consideration should be taken in this field.
Thus, improvement measures such as disabling one of the units, decreasing the HRT, optimizing the sludge
blanket depth, increasing loading rates and adding chemicals are currently being contemplated, also taking into
account the effects on the subsequent digestion process.
Figure 5

VIEW LARGEDOWNLOAD SLIDE

Performance indices for the secondary clarifier unit.
Figure 6
 VIEW LARGEDOWNLOAD SLIDE
Performance indices for the sludge thickening unit.
Figure 7

VIEW LARGEDOWNLOAD SLIDE

Performance indices for the anaerobic digestion unit (2016–2018).
The PX associated with the anaerobic digestor is shown in Figure 7(a). All present acceptable to good
performance except for the PX associated with the VFA/alkalinity ratio. Since alkalinity levels are consistently
within the typical range (Figure 7(a)), special care must be taken to maintain the VFA concentration at
acceptable levels. The reference values for VFA concentration should be defined for each WWTP as it will vary
with the sludge quality. The sludge quality will affect the bioavailability of the substrate and/or the
thermodynamic conditions (Carvalheira et al. 2018). Analysing Figure 7(b), it seems that this unit could operate
best by maintaining the VFA concentration above 0.4 g/L, taking care to maintain the other PX within acceptable
levels. Further detailed studies should be carried out to fully understand how to optimize this process as it
could potentially increase the biogas production efficiency, improving the indicator BP18 (Figure 2).

CONCLUSIONS
Listen
The studied WWTP presents an overall good performance in terms of effectiveness and reliability. However,
improvement measures are needed in terms of sludge management and energy performance. To understand
how to improve the energy performance, energy measurement campaigns were suggested and are currently
being performed to understand which equipment is underperforming. In addition, a more in-depth critical
analysis to the activated sludge unit will be performed to pinpoint improvement opportunities.
In terms of operation, this study showed that the oversizing of the WWTP is resulting in lower efficiency. To
overcome this, operational improvements were identified for sludge management, particularly in the primary
and secondary clarifier, the gravitational sludge thickening and anaerobic digestion unit.
The application of the PAS to the WWTP allowed the identification of improvement measures in both liquid and
solid treatment stages. Besides supporting the facility's operational management, this system provides
meaningful information that is used for decision-making purposes in terms of the utility's asset management
system by identifying the most critical assets that require intervention, rehabilitation or upgrade (equipment or
technology) as part of an integrated and on-going process that aims to continuously improve the facilities'
efficiency and effectivenes

ZIGOOOO

Outline
1. Highlights
2. Abstract
3. Keywords
4. 1. Introduction
5. 2. Overview of benchmarking methods for WWTP energy efficiency
6. 3. Comparison of benchmarking methods
7. 4. Actor analysis in WWTPs
8. 5. Proposed framework for energy efficiency improvement
9. 6. Conclusions
10. Declaration of competing interest
11. Acknowledgements
12. Data availability
13. References
Show full outline
Figures (6)
1.

2.
3.

4.

5.

6.
Tables (3)
1. Table 1
2. Table 2
3. Table 3

Journal of Environmental Management

Volume 344, 15 October 2023, 118624

Review
Energy efficiency in wastewater treatment
plants: A framework for benchmarking
method selection and application
Author links open overlay panelStefano Longo , Almudena Hospido, Miguel Mauricio-Iglesias
1

Show more
Add to Mendeley
Share
Cite
https://doi.org/10.1016/j.jenvman.2023.118624Get rights and content
Under a Creative Commons license
open access

Highlights
  •
Operation data are recorded and stored in WWTP but not fully utilized.

 •
Benchmarking methods can transform data into energy efficiency WWTP
diagnosis.

 •
Inefficiency can be decomposed and traced back to different root-causes.

 •
Benchmarking methods are linked with user's goals, skills, and data availability.

Abstract
Utilities produce and store vast amount of data related to urban
wastewater management. Not yet fully exploited, proper data analysis would provide
relevant process information and represents a great opportunity to improve the process
performance. In the last years, several statistical tools and benchmarking methods that
can extract useful information from data have been described to analyse wastewater
treatment plant (WWTP) energy efficiency. Improving energy efficiency at WWTPs is
however a complex task which involves several actors (both internal and external to the
water utility), requires an exchange of different types of information which can be
analysed by a broad selection of methods. Benchmarking method therefore must not
only be selected based on whether they provide a clear identification of inefficient
processes; it must also match the available data and the skills of those performing the
assessment and objectives of stakeholders interpreting the results. Here, we have
identified the requirements of the most common benchmarking methods in terms of
data, resources, complexity of use, and information provided. To do that, inefficiency is
decomposed so that the analyst, considering the objective of the study and the available
data, can link each element to the appropriate method for quantification and
benchmarking, and relate inefficiency components with root-causes in wastewater
treatment. Finally, a framework for selecting the most suitable benchmarking method to
improve energy efficiency in WWTPs is proposed to assist water sector stakeholders. By
offering guidelines on how integrates and links data, methods and actors in the water
sector, the outcomes of this article are expected to move WWTPs towards increasing
energy efficiency.

 Previous article in issue

 Next article in issue

Keywords
Data envelopment analysis
Regression analysis
Stochastic frontier analysis
Wastewater treatment
Key performance indicator
Actor analysis

1. Introduction
Thanks to new developments in the field of information technology, cheaper sensors,
and increasingly common supervisory control and data acquisition system (i.e. SCADA
systems), there is a vast amount of data on urban wastewater management, which has
been not fully exploited yet and constitutes a resource to improve the process
performance. Some of these data are collected and held by water management actors
including utility operators and different level of environmental authorities for different
purposes, without a common format or storage method. Overall, large amounts of data
from wastewater treatment plants (WWTPs) are being generated which need to be
properly transformed into knowledge for enhancing their operation (Corominas et al.,
2018).
In current practice, plant operators often have an overwhelming amount of data at their
hands, which are very difficult to process and analyse in a timely manner. Methods and
tools that enable systematic extraction of information from data sets would assist in
optimising the plant, eventually helping to further increase the effluent quality, to
reduce the consumption of energy and other resources and to foster the operator's
knowledge on the plant processes (Yoo et al., 2008). Out of the many processes through
which data can be transformed in knowledge such as classification, clustering,
prediction, neural networks, machine learning (Corominas et al., 2018), this work/paper
focuses on benchmarking techniques as a primary strategy for data management in
WWTPs and its application to the evaluation of energy efficiency.
As WWTPs are large energy consumers, energy efficiency is relevant for virtually
all water utilities. Besides, current regulation imposes energy efficiency audits in water
utilities larger than a certain size. Indeed, Directive 2012/27/EU on energy efficiency
(European Commission, 2012) specifies, through its national transpositions, that
European large companies must be audited in terms of energy efficiency. However, the
Directive does not specify important elements such as a clear definition of energy
efficiency for WWTP (Longo et al., 2019) or the methods to be used in the compulsory
energy efficiency audits. As discussed later, benchmarking methods can be suitable for
energy efficiency provided that the right method is chosen, and this selection must not
only be based on whether they provide a clear identification of inefficient processes but
also on the proper match between the available data and the skills as well as objectives
of those performing the assessment.
Energy benchmarking concerns a variety of people, from plant operators, managers,
regulators, technology providers, consumers, etc., with different capabilities and
interests which should be reflected on the election of the right tool (Walker et al., 2021).
A common characteristic of the methods developed for wide standardised
benchmarking is their simplicity, as they are intended for diverse actors. Examples of
such methods include the Energy Check developed by DWA in Germany (DWA -
German Association for Water Wastewater and Waste, 2015), the Energy Star sponsored
by the Awwa Research Foundation in the USA (Carlson and Walburger, 2007) or the
Standard CEN/TR 17614 developed with the support of the H2020 framework
programme (Longo et al., 2019). The International Benchmarking Network for Water
and Sanitation Utilities (IBNET, 2022), a World Bank sponsored initiative, counts a
large number of indicators for benchmarking which, in the case of WWTP energy
efficiency is based on a single indicator: the energy used per unit volume of wastewater,
usually reported in kWh/m (Walker et al., 2021). Differently, more detailed analyses
3

(based on data envelopment analysis, regression analysis, etc.) by technical actors

require more sophisticated and data-intensive tools.
First, a detailed overview of the methods is given in section 2 which may be skipped for
the experienced reader. Section 3 presents a comparison of the methods and how they
can provide a decomposition of energy inefficiency. Section 4 is devoted to establishing
which method is suitable for each actor. Lastly, section 5 presents an integrated
framework that links data, methods and the main actors involved, describing the flow of
information leading to decision making from data and hence into energy efficiency
investments. It is concluded highlighting the importance of method selection by each
actor for an effective promotion and communication of energy efficiency information in
the wastewater sector.

2. Overview of benchmarking methods for WWTP energy

efficiency
Efficiency is commonly defined as “the relationship between the production of a service,
good or energy” and its resource demand (European Commission, 2012). In the case of
energy efficiency, the resource is the energy demand, including other resources that can
be ultimately converted into a useful form of energy. Regarding WWTPs, defining the
service delivered is somewhat more complex. “Cleaning wastewater” or “recovering
resources” are not readily quantifiable and must therefore be decomposed in the
different functions of a WWTP which may include, among others, “eliminate nitrogen”,
“reduce the number of pathogens”, “remove phosphorus from the liquid streams”,
“produce biogas from organic carbon”, etc. depending on the influent, location
of discharge point and the requirements of the effluent. Therefore, the first step in
energy benchmarking is specifying the function(s) provided by the WWTP and, then,
relate them with the energy demand. As covered in detail in the following subsections,
all the benchmarking methods require this definition of energy (or “inputs”) and
services provided (or “outputs”).

2.1. Key Performance Indicators and ratios

Directly based on the definition of efficiency, Key Performance Indicators (KPI) are the
simplest benchmarking method. A KPI is often a ratio of an input and an output, which
can be obtained by normalizing the energy use based on the unit activity or service
provided. Based on the definition of KPIs, several WWTPs can then be compared using
input-output data (Fig. 1).
 1. Download : Download high-res image (90KB)
2. Download : Download full-size image
Fig. 1. Input-output combinations for several WWTPs. Dotted line and dashed line
indicate the best performance in nitrogen and COD removal, respectively.
Regarding COD, the plant with highest efficiency is the one with the lowest ratio of input
and output (i.e. WWTP 1), and its efficiency level is shown as the dashed line (Fig. 1).
For WWTPs, which usually operate close to the discharge limits, the inefficiency would
be the vertical distance to the dashed line, i.e., the energy that can be saved while
removing the same amount of COD.
The specific energy consumption has been used to roughly characterise the energy
efficiency of a WWTP and is reported in large datasets such as Mizuta and Shimada
(2010, 985 WWTPs in Japan), Yang et al. (2010, 599 WWTPs in China), Krampe (2013,
24 WWTPs in Australia), Vaccari et al. (2018, 241 WWTPs in Italy), Ganora et al. (2019,
300 WWTPs in Europe), Luo et al. (2019, 2022 WWTPs in China). However, a direct
comparison based only on the specific energy consumption misses indeed many details
in the plant operation and makes comparison only possible when the WWTPs are very
similar, i.e. similar layout, technology, climate, influent, etc.
The obvious advantage of KPIs is that they are relatively inexpensive to obtain, and easy
to implement and understand. However, this approach is limited in scope as it involves
only partial evaluations. As an example, Panepinto et al. (2016) in a thorough study on
Turin WWTP energy efficiency, used four KPIs (energy per person equivalent, per m of 3

water, per kg of COD removed and per kg of N removed) as, in effect, one KPI may not
fully reflect the purpose of the plant. A WWTP could have multiple functions, e.g.
removing COD, nitrogen, phosphorus, and pathogens, or producing energy or material
like biogas and fertilizers. As an example, evaluating the efficiency of WWTPs in
removing COD and nitrogen would require two KPIs. WWTP 1, in Fig. 1, has high
efficiency in removing COD but low efficiency in removing nitrogen. Conversely, WWTP
2 is the most efficient in nitrogen removal but has a poor COD removal performance. As
both nitrogen and COD are important objectives, WWTP 3 could be preferred as it
performs well in both dimensions. In this case, weighting between the two plant's
objectives would be necessary.
As a conclusion, partial benchmarks such as simple ratios and KPIs often make
misleading comparisons and therefore other strategies should be recommended for
effective energy benchmarking beyond a first approximation. In this regard, a proper
measure of WWTP energy efficiency should reflect a multidimensional concept by
considering the different functions of the plant.

2.2. ENERWATER methodology

With the aim of representing the multifunctional nature of WWTP endeavour, a
methodology was developed in the framework of the ENERWATER Coordination
Support Action to systematically determine the energy efficiency of a particular WWTP.
2

As a result of the application of the methodology stems the Water Treatment Energy
Index (WTEI), which is a composite index stemming from the aggregation of several
KPIs (Mauricio-Iglesias et al., 2020).
The full methodology is described in detail elsewhere (Longo et al., 2019) and briefly
summarised here for completeness. The approach basically consists of 1) measuring
energy use and WWTP outputs in terms of flowrate treated, COD and nutrient
removal, sludge disposal and pathogen load reduction; 2) determining or estimating
KPIs that represent the efficiency of the different WWTP functions (Table 1) as carried
out in different stages; 3) expressing the KPIs in a single index (i.e. the WTEI) that can
be used for benchmarking energy efficiency of different WWTPs and 4) assigning a label
(A, B, C …) for ease of communication to a broad public.
Table 1. Definition of KPIs used in the Rapid Audit methodology.

Stage KPI KPI units Meaning/function

Stage 1 KPI1 kWh/m3 Energy for pumping wastewater through the WWTP
Stage 2 KPI2 kWh/kg TSproc Energy for removing solids in primary treatment
Stage 3 KPI3 kWh/kg TPErema Energy for removal of COD, nitrogen and phosphorus
Stage 4 KPI4 kWh/(logRed·m3) Energy for removal of pathogens in tertiary treatment
Stage 5 KPI5 kWh/kg TSproc Energy for treatment of sludge and removal of solids
a
Where the total pollution equivalent TPE = COD (kgCOD) + 20 TN (kgTN) + 100 TP (kgTP)
as defined by (Benedetti et al., 2008).

There are two versions of the methodology that differ mainly on the level of detail of the
information required:

 i)
The Rapid Audit (RA) methodology determines the WTEI for benchmarking
using routine energy measurement and effluent sampling, hence providing a
quick benchmarking;

 ii)
 The Decision Support (DS) methodology requires an intense sampling campaign
with analyses of influent and effluent of different plant sections. It provides not
only the WTEI but also a diagnosis of the inefficient spots in a plant allowing the
user to propose corrective actions.
The results of applying the RA and the DS methodologies are then represented as labels,
common in other energy efficiency evaluations, so that they can be easily understood as
well by a broad public (Fig. 2).

1. Download : Download high-res image (1004KB)

2. Download : Download full-size image
Fig. 2. Representation of energy efficiency information by the WTEI and the
ENERWATER methodology (from Longo et al. (2019)).
After having been subjected through the complete revision, update and voting process
by the members of the European Committee for Standardization (CEN) the
ENERWATER methodology became the core of the new standard CEN/TR
17614 Standard method for assessing and improving the energy efficiency
of wastewater treatment plans, approved in January 2021 as the European
standard for defining and measuring energy efficiency in wastewater treatment plants.
As a CEN standard can be subjected to further modifications, the remaining of this
discussion will refer to the ENERWATER methodology noting that it can be partially
applicable to the CEN/TR 17614.

2.3. German standard DWA-A 216

The German Water Association (DWA) has devoted a considerable effort in the last two
decades in the development of a standard for energy efficiency improvement at WWTPs
(Clos et al., 2020). The proposed guideline, the German standard DWA-A 216 (2015),
consists of a dual approach. A first simple assessment called Energy Check (EC) is based
on the comparison of aggregated energy consumption KPI (i.e. kWh/PE year) and
quickly screening out plants where energy waste is present, hence in need of a more in-
depth analysis. This second approach, which is called Energy Analysis (EA), consists of
breaking down the plant to the level of individual equipment and comparing the energy
consumption versus ideal performance values estimated at medium load for energy-
consuming equipment such as pumps, blowers, compressors, agitators, sludge heating
but also “minor” equipment such as rakes, scrapers, etc. As the EA requires a lot of
operational data as input for the requested calculations, the whole process of an EA may
last for several months. The amount of work depends on reliability of data, the number
of machines and processes to be evaluated and the number of identified actions. It can
vary from a few days for a simple pumping station to several weeks for a complex
wastewater treatment plant (WWTP). Usually a hydraulic engineer and an electrical
engineer shall be involved in the EA.

This approach is certainly data-intensive, but it is very well-suited for the mechanical
revision of the WWTP, proposals for equipment renewal and comparison. The main
drawback is, perhaps, that the whole-process information is not used, e.g. even though a
blower is operating optimally, over-aerating by setting the dissolved oxygen setpoint too
high is a decision that would turn the process inefficient while it would not be captured
by the equipment-by-equipment analysis.

2.4. Data envelopment analysis

Data Envelopment Analysis (DEA) is a multi-factor productivity analysis model used for
estimating the relative efficiencies of a homogeneous set of firms (or decision-making
units in the context of the DEA literature). In general, DEA applications can be classified
in single- and two-stage DEA, based on whether efficiency estimation is conducted in i)
one stage considering only input and output variables, or ii) in two stages including a
second regression procedure that takes into account for the possible effect of exogenous
influences on efficiency such as e.g. climate or technology.

2.4.1. Single-stage DEA

Intuitively, this deterministic technique can be understood as an extension of KPI
analysis for multiple inputs and multiple outputs, thereby representing an attractive tool
for performance assessment (Cook and Seiford, 2009). Using the example of COD and
nitrogen removal (Fig. 1), a simple illustration of how DEA estimates efficiency is given
in Fig. 3.

1. Download : Download high-res image (105KB)

2. Download : Download full-size image
Fig. 3. Efficiency as estimated by DEA for two outputs (COD and N removal) and one
input (energy consumption).
From the definition of efficiency, WWTPs are more efficient if they use less input for the
same output, and/or produce a higher output with the same input. We therefore identify
the line connecting WWTP E, D and C as the efficient frontier; no point on this frontier
line can improve in one of the efficiency dimensions (or performance dimensions)
without worsening in the other. The method's name comes from this property as the
frontier is said to “envelop” these points. The efficiency of WWTPs not on the frontier
line can be measured by evaluating the distance to the frontier line (Amaral et al., 2022).
For example, the inefficiency of A can be evaluated by the ratio OP/OA, where OA, the line
from zero to A, crosses the frontier line at P (Fig. 3). An increase in efficiency can be
achieved that is by reducing inputs or increasing outputs. Since amount of pollutants to
be removed by the plant is usually fixed by the effluent quality requirement and the
influent composition, in general the only reasonable improvement of a WWTP consists
in input minimization. In this case we talk of input-oriented DEA model, in contrast to
the output oriented where the objective is output maximization.
Formally, the efficiency of a set of WWTPs can be estimated by the CCR DEA (Charnes
et al., 1978). For p inputs, q outputs and � WWTPs, we can determine the input-
oriented efficiency of the data matrix of input and output vectors (�,�), by solving for
each observation the following constrained linear programming problem:
(1)min�,��subjectto���≥����≥���≥0.where the index k represents a given
observation, � is the matrix of inputs, � is the matrix of outputs, and � is vector of
weights given to each observation. Problem (1) can be interpreted as combining WWTPs
(by weights λ) to produce an output level at least equal to plant k (��≥��) and then
selecting the combination with the minimum input level (���≥�� for minimum �).
Solving the linear programming problem (1) � times generates the efficiency indices ��,
one for each WWTPs. WWTPs with efficiency scores ��<1 are inefficient, since they
would be capable of reducing their input(s) without affecting the amount of output(s).
Efficient WWTPs receive efficiency score ��=1.
Thanks to its capability to include several inputs and outputs, DEA has been very
popular in academic research on WWTP efficiency during the last years. For example,
studying a set of 77 Spanish WWTPs, Hernández-Sancho et al. (2011) showed that plant
size, COD removed and aeration were determinant in energy efficiency; Dong et al.
(2017) reached similar conclusions analysing 736 Chinese WWTPs, highlighting as well
the importance of the climate type thanks to the broadly distributed sample. With a
broader perspective, Amaral et al. (2022) studied 120 wastewater service providers (not
just WWTPs) in Portugal finding a general increase of efficiency from 2015 to 2019, in
particular related to energy consumption.
The rationale of DEA is that the plant manager knows best the preferred objectives for
the operation. Therefore, instead of comparing several indexes (i.e. KPIs) or estimating
the expected energy consumption based on the several outputs to be fulfilled, a plant
receives the best possible score regarding its performance in each of several plant
objectives.

2.4.2. Two-stage DEA

When applying DEA to WWTP benchmarking, several authors have highlighted the
importance of accounting for exogenous factors (in principle, any factor that is not
under the direct control of the management). Carvalho and Marques (2011) warned on
the danger of misattribution of efficiency (or inefficiency) to managerial decisions if
exogenous variables are not taken into account in the water sector, including indeed
wastewater. Due to the complexity of the analysis of exogenous variables, Longo et al.
(2018) proposed a systematic method to apply DEA specially tailored for the analysis of
energy efficiency in WWTPs (i.e. two-stage DEA). In this approach, firstly proposed
within the wastewater sector by Guerrini et al. (2017), the basic idea is to estimate
efficiency scores in the first stage considering only the space of inputs and outputs,
ignoring the exogenous factors. Then in the second stage, a bootstrap-based algorithm is
used to assess the impact of the exogenous factors and obtain valid and accurate
inference for bias correction of the efficiency estimates. The specification used for bias
correction being:
(2)��=�0+����������������+��������������������+���������������
���+����������+������+������+����������+��;where ������� is
categorical variable indicating the country location of the
plant, ��������� end �������� are categorical variables indicating the type of
technology used in the plant as respectively secondary and tertiary treatment, �� is
the natural logarithm of the load factor, �� is the natural logarithm of dilution
factor, ���� is the natural logarithm of the annual average outdoor temperature,
and � is the random term. The results obtained by Longo et al. (2018) confirm that
apart from the amount of pollutants removed during the wastewater treatment, there
exist a number of exogenous determinants of the plant energy demand that wastewater
operators cannot control, e.g. the temperature the plant is operated. Energy variation
due to those factors can be misattributed to changes in efficiency. Adjusting for the
effect of those factors can lead to substantial changes in efficiency estimates depending
on the adverse or favourable environmental conditions a WWTP is operating.

2.5. Parametric approaches

Parametric approaches to benchmarking in the contest of wastewater treatment is
relatively new in comparison to e.g. KPI and DEA approaches. The parametric approach
consists on statistical analysis that based on the technique can be divided in i)
simple regression analysis, e.g. Ordinary Least Squares (OLS) or Corrected OLS (COLS),
and ii) Stochastic Frontier Analysis (SFA).

2.5.1. Regression analysis

A linear regression model, such as the very popular ordinary least squares (OLS)
(Kutner et al., 2005), can be used to estimate the energy use of a system. If plant data
are available, parameters � can be found that relate energy use, E, and a matrix of
variables of plant and operation characteristic, �, for �=1…�:(3)��=���+��;�∼�(0,�2).
The dependent variable (E) is also called the response variable. Independent variables
(X) are also called explanatory or predictor variables. An OLS model provides the
expected average energy use for a plant given its operating characteristics such the
amount of pollutants removed. The result of such estimation, in the simplest case, can
be a regression function with just one variable (OLS in Fig. 4).

1. Download : Download high-res image (94KB)

2. Download : Download full-size image
Fig. 4. Comparison of the referenced performance: efficient frontiers (SFA, COLS) and
average performance (OLS).
In contrast with DEA, which establishes an efficient frontier (representing best
performance), the regression function estimated by Eq. (2) represents the average
performance. The residuals are defined as the distance from each data point to the
regression line, which can be positive or negative, if the data points lie, respectively,
above or below the line.
The application of OLS to benchmarking is based on comparing the performance of a
WWTP with the average performance. As the residual is the difference between the
actual energy use and the predicted energy use for a given service (or output), residuals
can be treated as measures of inefficiency. If the actual energy use of a given WWTP is
less than the predicted energy use (negative residual), it means that the WWTP uses less
energy than the average WWTP described by the regression line. Therefore, WWTPs
with ratings above the average can be considered inefficient while those with ratings
below are efficient.

A regression energy demand exponential function is developed in Niu et al. (2019). The
model is expressed as follows:(4)�=������������������������where � is
the electric energy consumption, � is the base of the natural logarithm, � is the flow
rate, � is the age of the plant facility, � is the wastewater collection area of the pipe
network, � is the amount of sludge produced and processed, ��� and ���� are COD
concentration in the influent and effluent, respectively; �, �, �, �, ��� and ���� are
all regression coefficients; and � captures the impact of categorical variables including
treatment technology, operating conditions, topography, and other regional variables.
The regression approach can be complemented with establishing a distribution of
residuals, which can assign a performance percentile to each WWTP. This approach was
followed by the Energy Star method (Carlson and Walburger, 2007), launched in
the USA in 2007 as part of a broader initiative of the Environmental Protection Agency
(EPA). This method is based on linear regression function developed using data of 257
WWTPs throughout the USA. The prediction of the average plant consumption is done
using as inputs the average influent flow rate, the influent and effluent BOD
concentration, the plant load factor and two binary variables accounting for whether the
plant presents filtration and/or nutrient removal. Then, the actual benchmarking score
can be obtained by comparing the difference between the actual consumption and the
prediction (the residual) with a distribution of residuals. Negative residual means that
the plant uses less energy than expected and vice-versa. Spruston et al. (2014) checked
the validity of the Energy Star method on 35 Canadian WWTPs concluding that it was a
valid method for energy benchmarking although it was not fully adapted to Canadian
specificities and for certain types of WWTPs.
A consequence of using a parametric approach such as OLS is that the residuals are
treated as a measure of inefficiency while they can partially represent variables not
included in the regression function. For example, in the Energy Star function, the
variation in energy use due to nutrient removal is modelled as a categorical variable
(either yes or no) when in reality it is the amount of nutrients removed that will impact
the energy demand. Then, much of the variation in efficiency scores may remain
unexplained by not including the right explanatory variable (the amount of nutrient
removed).

An extension of OLS which provides a demand frontier, instead of an average

performance line, is the Corrected Ordinary Least Squares (COLS). This method is a
two-stage procedure: i) the regression line is first estimated through OLS and ii) then
the regression line is shifted downwards such that the resulting regression line (called
the COLS frontier) envelops all data. Hence, all the residuals are positive, except for the
WWTPs that are found to be efficient (where the residual is zero). Therefore, COLS
assumes that all residuals are due to inefficiencies (Fig. 4).
While the COLS method has its appeal in terms of simplicity, a more realistic view is
that not all the differences between the actual data and the frontier are due to efficiency.
This is especially relevant in wastewater application where significant data uncertainty
may be present both in data gathered under either compulsory monitoring or reporting
requirements under law (Yoshida et al., 2014). Furthermore, as the COLS frontier is
meant to envelop all the points, it is very sensitive to outliers which can move the
frontier away from the regular data points.

2.5.2. Stochastic Frontier Analysis

The COLS approach implies that the residual is due solely to inefficiency. Alternatively,
the residual can be considered as composed by a random (stochastic) error term and an
inefficiency term, which may be composed itself of several contributions.
The econometric technique known as Stochastic Frontier Analysis (SFA) has been
developed to provide separate estimates of these two components. A general version of
the stochastic frontier model can be described as:(5)��=�(��;�)
+��;��=��+��,;��∼�(0,��2);��∼�+(0,��2),where � is energy use, � is a vector of variables
influencing energy use, including plant characteristics and external factors,
and � includes all the parameters to be estimated. Energy use residuals are now
captured by two random terms, � and �. The noise term ν in (5) is assumed to be
normally distributed while the non-negative random term �, which represents
inefficiency, is generally assumed to follow a half-normal (N ) or other positive
+

distributions. It is then possible to estimate parameters �, σ and σ using maximum

ν u

likelihood methods (Kumbhakar and Lovell, 2003). A comparison of the stochastic

frontier, the COLS and OLS models is provided in Fig. 4.
Environmental factors can be represented as a function of some observable variables but
many relevant aspects that cause heterogeneity are unobserved or not known a priori. If
panel data are available, that is, each unit is observed at several different points of time,
part of these unobserved heterogeneity can be estimated as it does not change with time
(Wooldridge, 2010). The simplest SFA model for panel data hence can be rewritten as:
(6)���=�(���;�)+���;���=���−��;��≥0;�=1,…�;�=1,…,�,where �(���;�) is a linear
function of the variables in the vector ���, index t indicates different time points
and ��≥0 is the energy term of plant �. The model in (6) can be estimated assuming
either �� is a fixed parameter (the fixed-effects model) or a random variable (the
random-effects model). This approach no requires distributional assumptions
on �� and is, thus, labelled as distribution-free (Schmidt and Sickles, 1984). Another
important advantage of using panel data over cross-section is that it is possible to think
of the inefficiency term as comprised by two components: persistent (i.e. time-invariant)
and transient (i.e. time-varying) (Filippini and Greene, 2016; Tsionas and Kumbhakar,
2014). The persistent component corresponds to the presence of structural problems
such as inefficient equipment or design limitations that prevent the plant from minimize
the use of energy; the transient component may be caused by the presence of non-
systematic difficulties that can be solved in the short term such as adaption of wrong
operational strategies due to e.g. too infrequent sampling.
To the best of our knowledge, the first application of SFA to efficiency in WWTP was
carried out by Longo et al. (2020), who studied panel data from 183 Swiss WWTPs
during 15 years. The WWTP energy demand function for panel data is described as
follows:
(7)ln���=�0+��ln��+�����ln������+����ln����+����ln�����+���4ln��4��+���
3ln��3��+�����ln������+∑�=16������������+���������+���−��−���where �
is energy consumption, � is the real price of energy, ���� is the volume of wastewater
treated, ��� is the plant capacity expressed as design flow rate, ���, ��4 and ��3 are
the pollutants concentration removed from wastewater, ���� is the average
temperature, ���� represents dummy variables to control for the effect of the type of
secondary treatment, ��� is a dummy indicating whether the plant also carries
out sludge dewatering.
In this specification, the random error term � is decomposed as ���=���−���,
where ��� is the inefficiency and ��� is statistical noise. The inefficiency part is further
decomposed as ���=��+��� where �� is the persistent component and ��� is the
transient component of inefficiency. The former is only plant-specific, while the latter is
both plant- and time-specific. In addition to that, SFA can be also extended to take into
account for unobserved heterogeneity and to include variables that are direct
3

determinants of efficiency level such as e.g. the age of the WWTP, as previously reported
by Castellet-Viciano et al. (2018) and Niu et al. (2019). Thanks to that, it is possible to
estimate inefficiency related to WWTP obsolescence and, therefore, how much energy
can be saved when a WWTP equipment is renewed.
The application of SFA to WWTPs whenever panel data is available is a novel and
promising approach able to distinguish whether inefficiency that originates in inefficient
equipment or from the inefficient use of (in)efficient equipment. Hence, SFA represents
a useful tool to deduce energy diagnosis from common operational data.

3. Comparison of benchmarking methods

From the description of the benchmarking methods is easy to infer that they differ in
complexity, scope, sensitivity to outliers, data requirement, efficiency interpretation etc.
as summarised in Table 2. It becomes clear that no method can provide a universal
solution to benchmarking, as all of them face their own problems both on the theoretical
and the practical side. This implies that the final efficiency estimates should not be
interpreted as being definitive measures of inefficiency. Rather, a range of efficiency
scores may be developed and act as a signalling driver rather than as a conclusive
statement.
Table 2. Summary of comparison of different benchmarking methods.

Empty Cell Advantages Disadvantages

KPIs  -  -
Simple and already used in the Being based on KPIs, it implicitly
wastewater treatment sector. assumes constant return to scale.

 -  -
Can be adapted for different Does not account for exogenous
Empty Cell Advantages Disadvantages
WWTPs and objectives. factors, beyond WWTP control.

 -  -
Minimal data requirements. Difficult interpretation for multiple
inputs/outputs.

 -
No diagnosis/corrective actions.
ENERWATER  -  -
A composite indicator can Being based on KPIs, it implicitly
summarize the multi-objective assumes constant return to scale.
purpose of a WWTP.
 -
 - The database on which the
It can be applied to different plant benchmarking system is developed
layouts. covers mostly WWTPs of medium
size and located in Europe.
 -
Being equivalent to percentile it is  -
easily interpretable. Composite indicators cannot directly
account for exogenous factors,
 - beyond the WWTP control.
Rapid Audit version requires
simple and common routine data  -
normally available in all water Decision Support requires detailed
utility. energy monitoring and sampling
campaign.
 -
Decision Support can be used to  -
do diagnosis of inefficient Not ready for multiple outputs
stages/functions of a WWTP. (ready though for biogas
production).
Standard  -  -
DWA-A 216 Energy Check version used Energy Check version is limited to
common and widely KPI plants with similar characteristics
(kWh/PE). (not applicable to other countries).

 -  -
Energy Analysis version is very Energy Analysis version is very
useful for analysing single data-intensive (i.e. equipment
equipment energy consumption. technical sheets are required) and
time requiring.
OLS/COLS  -  -
OLS is intuitive and widely used. It provides an average performance
as benchmark (OLS).
 -
Easy to account for exogenous  -
factors. Extremely sensitive to outliers and
Empty Cell Advantages Disadvantages
 - measurement error (COLS).
A distribution function of the
residuals can be used to assign  -
performance percentiles. Does not lead to diagnosis or
corrective actions.

 -
Cannot be used for multiple outputs.
DEA  -  -
DEA is easy to extend to multiple Corrective measures are difficult to
outputs and inputs. identify from DEA outcomes.

 -  -
DEA can be implemented on a Efficiency scores are sensitive to the
relatively small dataset in choice of input and output variables.
comparison to regression analysis.
 -
 - It is very sensitive to outliers which
Once the input and output can move the frontier away.
variables have been selected, DEA
is quick and straightforward to  -
implement. For DEA, testing a new WWTP
requires solving the model again for
 - the whole set of observations.
It doesn't require assumptions
about the frontier function.  -
As more inputs and outputs are
 - considered, the number of WWTPs
Two stage DEA can account for on the frontier increases and the
factors that are beyond the control discrimination power decreases
of the WWTPs. accordingly.
SFA  -  -
SFA distinguishes persistent and High data requirements. Panel data
transient inefficiency; thus, it can required to distinguish between
be used as a diagnosis tool. unobserved heterogeneity and
inefficiency, as well as persistent
 - and transient inefficiency.
Using panel data, it can take into
account unobserved/unmeasured  -
factors. Requires a priori assumptions, e.g.
specifying a functional form and
 - statistical distributions for the
SFA allows statistical inference inefficiency terms.
about which parameters to include
in the frontier estimation.  -
Can be difficult to implement due to
 - the complexity of algorithm required
Environmental variables can be
directly included in the regression  -
Empty Cell Advantages Disadvantages
model as regressors. Cannot be used for multiple outputs.

 -
SFA incorporates the possibility
of separate measurement error and
stochastic factors, being robust to
outliers.
A relevant point highlighted in the method characterisation and comparison reported
in Table 2 is the capability to handle exogenous factors, understood as any factor that is
not under the direct control of the management, therefore exogenous to the WWTP
system, such as influent characteristics. Labelling a factor as exogenous might, however,
be a debatable issue and in fact there is certain controversy around which factors are
legitimate uncontrollable influences on performance and which ones should and/or can
be controlled and by whom. A factor can be uncontrollable for one stakeholder but not
4

for another. An example is the plant size. Inefficiency due to scale is exogenous to the
plant operator but may be endogenous for the water utility management, who may
decide to merge two close plants in order to operate only one bigger plant. Furthermore,
there may be factors, such as the load factor, that depending on the context may have
different interpretations. Oversizing may be a solution to seasonal variation in the load
entering the plant in case of e.g. a plant operating in a touristic area but may also be due
to erroneous design estimation. In this case, inefficiency due to oversizing is beyond the
control of an operator but can be eliminated by a water utility by e.g. dividing the plant
in different treatment lines in order to modulate the load. Furthermore, since energy
efficiency is only a secondary, although important, objective for water utilities that
should never jeopardize the primary objective to clean water, it may be preferred higher
energy consumption and a more robust WWTP. In practice in the short run, very little
may be controllable by plant operators, whilst in the longer-term inefficiency due to
factors such as plant size or load factor can be solved.
The main difference between one method and another is how each method highlights
differences in energy consumption between DMUs, i.e. inefficiencies. Comparing
efficiency scores for different DMUs within methods provides better indications than
comparing efficiency scores inter methods. In this regard, relating the inefficiencies
between the different methods can be particularly useful as it can provide a plausible
framework for interpretation and driving decisions, which otherwise would be difficult
when using just benchmarking method. For example, despite the popularity in academic
circles of tools such as DEA, due to its ability to aggregate multiple inputs and outputs in
a single efficiency measure, it is plausible that, from a managerial point of view, this is a
weakness, as it distracts attention from the question of where the problems actually lie
and where one should search for ideas for improvement.

Before thinking about solutions to energy efficiency problems, the logical first step is to
diagnose where the problem, i.e. the inefficiency, is located. Inefficiencies may come
from factors beyond the control of the management or internally e.g. due to old and
inefficient equipment or due to an inefficient use of the equipment itself. Within the
plant boundaries, plant managers may discover that a plant provides aeration with an
inefficient blower or diffusers. Alternatively, they may discover efficiency problems
through internal monitoring of their own performance. In this context, efficiency
metrics that treat the WWTP as a black box have limited utility as they do not pinpoint
where to target the intervention. Ultimately, in order to identify inefficiencies and, as a
result, proper improvement strategies, the analyst must be aware of the specific
methods which are helpful to identify a particular type of inefficiency. To do that,
inefficiency needs to be decomposed (Fig. 5) so that the analyst, considering the
objective of the study and the available data, can link each element to the appropriate
method for efficiency quantification and benchmarking.

1. Download : Download high-res image (285KB)

2. Download : Download full-size image
Fig. 5. Inefficiency decomposition by different methods.
The inefficiency decomposition presented in Fig. 5 is then particularly useful for
defining better improvement strategies and diagnosis of inefficiency. In fact, the
application of such mapping of efficiency is fundamental to understand what type of
inefficiency a plant is affected and therefore to design specific improvement strategies. If
one does not pinpoint the fundamental problem, corrective actions and improvements
are not possible. Inefficiency need to be understood and a consensus on where exactly it
comes from need to be reached before to move forward. With the problem identified and
the data collected to substantiate it, the problem can be traced to its various causes.
With this analysis completed, it can be prioritized which problems should be addressed
promptly and in what order considering their costs.

4. Actor analysis in WWTPs

Many actors are involved in the wastewater treatment sector, from WWTP operators to
CEOs, including researchers or policy makers, among others, with different working
environment, type of decisions to confront or to take. Unlike academic researchers,
operators typically spend a significant part of their time facing unexpected and urgent
problems. In contrast to policy makers, plant operators have relatively limited and weak
levers for driving and securing change as their main objective is to guarantee the
operation of the plant within the existing regulatory framework, and communicate with
stakeholders at different levels, including institutions. In short, with very limited time
and capacity, operators must take informed decisions based on available knowledge.
Selecting a benchmarking method is indeed a complex task that this section tries to
address by answering the following questions:

 •
Who is performing the benchmarking (“actor”)?

 •
What is the goal of the benchmarking (“task”)?

 •
What method is best suited for both the task and the actor?

 •
What is the nature of the available data?