Wat. Res. Vol. 35, No. 11, pp.

2702–2710, 2001
# 2001 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: S0043-1354(00)00553-4 0043-1354/01/$ - see front matter

NUTRIENT ADDITION TO ENHANCE BIOLOGICAL

TREATMENT OF GREYWATER
BRUCE JEFFERSON, JOANNA E. BURGESS, AUDE PICHON,
JOANNE HARKNESS and SIMON J. JUDD*
School of Water Sciences, Cranfield University, Cranfield, Bedford MK43 0AL, UK

(First received 12 July 2000; accepted in revised form 1 November 2000)

Abstract}This study compares the chemical oxygen demand (COD) removal and respiration rates of a
microbial population treating real and synthetic greywaters dosed with nutrient supplements. The nutrient
composition of the real and synthetic greywaters was analysed and the dosing regime for nitrogen,
phosphorus and a range of trace metals planned accordingly. The doses consisted of eight single additives
(macronutrients and trace metals) to the control greywater and six trace metal additions to C : N : P
balanced greywater. The COD removal for the control real and synthetic greywater in lab-scale activated
sludge systems (0.038 and 0.286 kg COD kg MLSS
1 d
1, respectively) confirmed nutrient limitation and
the poor degree of greywater treatment. Nutrient dosing increased the COD removal rate and oxygen
uptake rate in many cases. The greatest stimulation of microbial activity was observed with zinc additions
to C : N : P balanced real greywater (1.291 kg COD kg MLSS
1 d
1 over 30 times the control). Inhibitory
effects to various extents were rare and limited mainly to the additions of metals to synthetic greywater.
The dominance of chemicals effects was observed on addition of some micronutrients; notably iron and
aluminium, metals on which many coagulants for use in biotreatment of other wastewaters are based. The
data indicate that the impact of understanding microbial processes and the nutrients required for
wastewater treatment can only serve to optimise process efficiency for the proposed treatment of
greywater. # 2001 Elsevier Science Ltd. All rights reserved

Key words}greywater, nutrient addition, microorganisms, micronutrients, trace metals

INTRODUCTION large-scale recycling at the Millennium Dome (Jef-

The water supply–demand balance in the UK is
projected to reach critical levels in the future due to
the anticipated growth in new housing developments
coupled with an increased per capita water usage and
decreasing supply in some parts of the country.
Water recycling forms one key component of the
strategy to offset this imbalance. In the short to
medium term, domestic recycling has been identified
as one of the most promising options. Urban water
use is higher in the UK than in many other European
countries, and is significantly more condensed into
urban centres; 50% of the population live in towns of
100,000 or greater population (Laine et al., 1999).
The focus of domestic water recycling has been
towards greywatery reuse for toilet flushing or
recreational irrigation, offering a potential 30%
reduction in potable water use. This has been
implemented at a number of sites around the UK
ranging from reuse within a single occupancy to
*Author to whom all correspondence should be addressed.
Tel.: +44-1234-750111 ext. 2542; fax: +44-1234-
751671; e-mail: s.j.judd@cranfield.ac.uk
y
Wastewater from all domestic sources except toilet
flushing.
2702
ferson and Judd, 1999). The current trend has been to
include biological treatment in the processing of
greywater for all but the very small scale, to restrict
the propensity for regrowth of biomass in post-
treatment water. However, unlike sewage, little is
known about the fundamentals of greywater treat-
ment and hence its suitability for biological treat-
ment. Experience from pilot trials with biological
systems treating greywaters has shown that such
systems are generally limited to a low organic load,
while analysis of the feed waters has shown a high
COD : BOD ratio coupled with nitrogen and phos-
phorus deficiencies (Laine et al., 1999). It has been
hypothesised that addition of macro- and trace
nutrient addition may overcome both these pro-
blems, thus increasing the efficacy of biological
technologies for recycling applications.
Biological cells require six macronutrients for
metabolic processes: carbon, oxygen, hydrogen,
nitrogen, sulphur and phosphorus. The primary
nutrients are nitrogen, phosphorus and carbon
(Beardsley and Coffey, 1985). The trace element
requirements of bacterial cells can be determined
from the composition of the cells (Metcalf and Eddy,
1991). However, as excess micronutrients can be
 Nutrient addition in greywater 2703

Table 1. Microbial nutrient requirements and the concentrations present in greywater

Nutrient Reported Role of trace metala Concentration detected (mg l
1)

requirements
(mg l
1)a
Real greywater Synthetic
greywater

N, P, S 15, 3, 1 minb } 5.00, 1.37, 16.3 5.00, 0.047, 17.5

Ca 0.4–1.4 Cell transport systems and osmotic balance in all bacteria. 47.9 47.0
Bridging anionic ECP and aiding flocculation. Increase growth
rates. Requirements and effects vary. Interacts with other metals.
K 0.8–>3.0 Cell transport systems and osmotic balance in bacteria. 5.79 3.96
Fe 0.1–0.4 Growth factor in bacteria, fungi and algae. Adsorbed in 0.017 0.009
proportion to the concentration available. Electron transport in
cytochromes. Synthesis of catalase, peroxidase and aconitase. Ion
reduction for floc formation.
Mg 0.4–5.0 Enzyme activator for a number of kinases and phosphotransferase 5.29 5.02
in heterotrophic bacteria.
Mn 0.01–0.5 Activates bacterial enzymes. Often interchangeable with 0.04 0.02
magnesium in kinase reactions. Lower affinity for binding sites
than other metals but still can inhibit metabolism at 1 mg l
1.
Cu 0.01–0.5 Bacterial enzyme activator required in trace quantities. Can inhibit 0.006 0
metabolism. Chelates other substances, reducing their toxicity.
Al 0.01–0.5 Not known. Affects the species found in sludge (Annaka, 1977). 0.003 0
Zn 0.1–0.5 Bacterial metallic enzyme activator of carbonic anhydrase and 0.03 0
carboxypeptidase A. Dissociable on active site of enzymes.
Stimulates cell growth. Toxic at 1 mg l
1, especially to protozoa.
Can exacerbate toxic effects of other metals and inhibit
metabolism.
Mo 0.2–0.5 Molybdenum is a common limiting nutrient (Grau, 1991). 0 0
Co 0.1–5.0c Bacterial metallic enzyme activator. Dissociable on active site of 0 0
enzymes. Activates carboxypeptidase for synthesis of vitamin B12
(cyanocobalamin) but otherwise toxic. Can inhibit metabolism.
a
Burgess et al., (1999a).
b
From COD : N : P ratio, Beardsley and Coffey (1985).
c
Sathyanarayana and Srinath (1961).

adsorbed onto cell walls, the concentrations in
biomass ash may exceed the actual required amount
(Wood and Tchobanoglous, 1975; Nicholas, 1963).
Actual requirements are influenced strongly by the
organic and hydraulic loading rates on the activated
sludge microorganisms, the cell growth rate (Speitel
and Segar, 1995), the nature of the wastewater, and
the mean cell residence time in the aeration basin
(Wood and Tchobanoglous, 1975).
As environmental conditions and nutrient supply
approach optimum the hydraulic retention time
necessary for wastewater treatment decreases (Speitel
and Digiano, 1988; Benefield et al., 1979). The ratio
at which nutrients should be supplied is contentious,
with COD : N : P ratios of 100 : 20 : 1 (Metcalf and
Eddy, 1991), 250 : 7 : 1 (Franta et al., 1994) and
100 : 10 : 1 and trace sulphur (Beardsley and Coffey,
1985) quoted in the literature.
In the cases where treatment plants are dealing
with a nutrient-limited wastewater, supplementation
with macronutrients can result in an enhanced
degradation of organics (Valo et al., 1985; Singleton,
1994). Adequate trace nutrient concentrations, in-
cluding metal ions and vitamins, are required to
support all the genera present in the activated sludge
treatment system in order for a diverse community to
survive. The lack of specific nutrients can result in an
unbalanced mixed culture, leading to a poor-quality
effluent, reduced treatment efficiency and sludge-
handling problems (Pala and Sponza, 1996; Soddell
and Seviour, 1996; Blackall et al., 1996). A clear
example of this can be seen with activated sludge,
where pin point flocs, dispersed growth and poor
flocculation have been attributed to deficiencies in
trace elements and have been shown to be amelio-
rated by nutrient dosing (Gostick, 1991). Therefore
the application of biological processes to the treat-
ment of greywater, and its subsequent optimisation is
a logical progression. The aim of this study was to
investigate the feasibility of optimising the biological
treatment of greywater through nutrient supple-
mentation.
MATERIALS AND METHODS
Atomic emission spectrophotometry using an inductively
coupled plasma atomic emission spectrophotometer (Ther-
mo Jarrell Ash, Atomscan 16) was used to identify the
missing or deficient trace elements in synthetic and real
greywaters as compared to theoretical requirements of
micro-organisms. The concentrations of available nitrogen,
phosphorus and sulphur were also measured and compared
with the theoretical requirements of bacteria for these
nutrients stated in the literature. This procedure produced a
list of nutrients required as a supplementation dose
(Table 1).
A series of tests was performed using a closed-cell aerobic
respirometer (CES Ltd. Aerobic Respirometer). This model
2704 Bruce Jefferson et al.

operates in a similar way as the Warburg respirometer
(Heddle, 1980). The respiration rate is the rate at which
oxygen is used as the terminal electron acceptor, and
therefore can be used to indicate the condition of the sludge
in terms of metabolic activity. Six trace elements with and
without nitrogen and phosphorus balancing of the grey-
water, as identified in literature (Table 1), were tested in
single supplements to both real and synthetic greywaters
(Table 2). The real greywater was collected from a purpose-
built greywater collection facility fed from 16 residential
flats in the University’s halls of residence. The water was
collected from the flats’ baths, showers and handbasins
only. The synthetic greywater recipe (Table 2) was devel-
oped from a recipe used in the UK water industry which
represents a normalisation of the variation of real grey-
waters found around the UK. Part of the remit of this study
was to evaluate the synthetic greywater recipe for differences
in characteristics between the two types of greywater and
assess the suitability of the recipe for use as a real greywater
substitute.
The sludge was taken from two membrane bioreactors,
one treating real greywater (COD : N : P of 13 : 4 : 1) and
the other treating synthetic greywater (COD : N : P of
894 : 106 : 1). The sludge was maintained in porous pots
(Water Research Centre, 1978; USEPA, 1996) at a pH of
7.5  0.2. Aliquots of the activated sludge (50 ml) were also
taken from the same membrane bioreactors, at the same
time and placed in glass respirometer cells to which
predetermined quantities of nutrient supplements had
already been added (Table 3). Control samples receiving
deionised water in place of supplements were also tested
when the test cells were dosed with micronutrients alone,
and controls receiving just nitrogen and phosphorus were
used when the test cells received nitrogen, phosphorus and
micro-nutrient doses. The volumes in all cells were made up
to 55 cm3 with deionised water as required to standardise
the partial pressures within the cells and make the tests
comparable. The respirometer cells were stabilised at 208C
and stirred continually for 15 h, the maximum time possible
before oxygen in the headspace of the cells became
exhausted and the biomass placed under oxygen-limited
conditions. Oxygen uptake was recorded automatically at
5 min intervals. A sample of the original activated sludge
was filtered for influent COD (Whatman GF/C filter) and
Table 2. Recipe for synthetic greywater
measurements of effluent COD were made at the end of each
run in the same way (Hach system, Camlab, adapted from
APHA, 1992). The mixed liquor suspended solids (MLSS)
concentration of the sludge was determined using standard
methods (APHA, 1992). Tests were duplicated to assess
repeatability.
RESULTS
The inductively coupled plasma (ICP) atomic
emission spectrophotometer analysis showed that
the greywaters were deficient in nitrogen, phosphorus
and a number of micro-nutrients, namely iron,
copper, aluminium, zinc, molybdenum and cobalt
(Table 1). The effects of nutrients on biomass range
from stimulatory to inhibitory:
* Increased COD removal associated with increased
or unchanged oxygen uptake indicates stimula-
tion.
* Decreased COD removal and oxygen uptake
indicates metabolic inhibition.
* Decreased COD removal and increased oxygen
uptake indicates the uncoupling of metabolism, in
which anabolism is inhibited and catabolism is
not.
* Increased COD removal and decreased oxygen
uptake may indicate metabolic uncoupling and
the inhibition of catabolism, but a more probable
mechanism is increased adsorption of COD-
imparting substrate components onto the bacter-
ial cell walls.
As the testing comprised several experiments, the
data were normalised so that the control results
equalled 100%, and the test results could be reported
as a percentage of the concurrent control, thus
making individual tests directly comparable (Tables
4–6, Figs 1 and 2).

The following are mixed and made up to 10.0 l with tap water
Nitrogen/phosphorus-limited greywater
0.64 g Synthetic soap Addition of nitrogen or phosphorus to real grey-
8.0 ml Hair shampoo
0.1 ml Sunflower oil water resulted in stimulation of the biomass (Fig. 1).
24.0 ml Tertiary effluent The addition of nitrogen to synthetic greywater
had the expected stimulatory effects on the sludge;
normalised oxygen uptake was 106% and COD
Table 3. Nutrient additions to the greywater samples removal 150% (Table 4, Fig. 2). Additions of cobalt,
Nutrient Dose added to real Dose added to synthetic
copper, iron and molybdenum to real greywater also
greywater (mg l
1) greywater (mg l
1) stimulated the biomass, resulting in normalised COD
removals of up to 200% with molybdenum (Table 4,
N 10.00 10.00 Fig. 1). Addition of micronutrients to synthetic grey-
P 1.63 2.95
S None None water led to stimulation of biomass by copper,
Ca None None molybdenum, zinc and aluminium. The COD re-
K None None movals achieved were 124, 117, 132 and 115%. Real
Fe 0.39 0.39
Mg None None greywater dosed with zinc and aluminium also
Mn None None showed chemical effects, as normalised COD removal
Cu 0.04 0.05
Al 0.05 0.05
increased to 154 and 225%, and oxygen uptake
Zn 0.97 1.00 decreased to 97 and 94%, respectively (Table 4,
Mo 0.70 0.70 Fig. 1). In synthetic greywater (Table 4, Fig. 2), the
Co 5.00 5.00
addition of cobalt, iron and phosphorus resulted in
 Nutrient addition in greywater 2705

decreases in oxygen uptake to 95, 81 and 88%, while

Repeatability (%)
the COD removal increased to 135, 103 and 125% of

0.3
3.2

2.2
3.1
2.1

1.1

1.1
1.7
1.6

5.7
3.8
4.3

23.3

26.3
the control, respectively.

COD removal
Nitrogen/phosphorus-balanced greywater
Dosing cobalt, copper, molybdenum and zinc into
Rate (kg kg MLSS d
1)

nitrogen/phosphorus-balanced real greywater stimu-

lated the biomass; COD removal increased to 130,
0.8
0.7
1.0
1.3
0.8
0.7
0.4
0.3
0.7
0.4
0.5
0.5
0.6
0.7
240, 150 and 290% of the control, respectively
(Table 4, Fig. 1). Dosing iron and molybdenum into
Synthetic greywater

synthetic greywater was also stimulatory, producing

normalised COD removals of 115 and 150%,
respectively. Adding aluminium to real greywater
Repeatability (%)

led to inhibition of the metabolism of the biomass;

COD removal dropped to 85% of the control, while

11.5

15.5
16.5

16.3
4.2

2.4

3.1
2.7

4.4
2.8
5.8
1.2
4.0

7.5
oxygen uptake rates remained the same (Table 4,
Fig. 1). Aluminium was also inhibitory in synthetic
greywater, as was zinc; both micronutrients caused
Oxygen uptake

the COD removal to decrease to approximately 80%

1
d )

of the control, and normalised oxygen uptake fell to

Table 4. Respiration and COD removal rates of activated sludge

1

70% for aluminium and 79% for zinc (Table 4,

Rate (kg kg MLSS

Fig. 2). Dosing iron into the real greywater resulted

1.7
1.4
2.2
2.5
1.6
1.5
0.8
0.6
1.0
1.0
1.0
1.0
1.2
1.4

in increased COD removal (185%) and a slight

decrease in respiration (97%).

Effects of greywater source and nitrogen/phosphorus

Repeatability (%)

balancing
Paired t-tests were applied to the respiration and
49.0
26.5
34.2
42.7
21.6
76.2

14.3
1.5

7.7

5.5
0.8
}

COD removal rates calculated per unit MLSS, and to

the normalised data. Paired t-tests perform a paired
two-sample Student’s t-test to determine whether a
COD removal

sample’s means are distinct. This t-test form does not

1
d )

assume that the variances of both populations are

1

equal. A paired test is appropriate when there is a

Rate (kg kg MLSS

natural pairing of observations in the samples, such

0.2
0.3
0.3
0.2
0.4
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.3
0.2

as when a sample group is tested twice}in this case,

testing micronutrients with and without nitrogen/
phosphorus balancing and with real and synthetic
Real greywater

greywater. Eight tests were carried out with con-

fidence limits of one standard deviation (Table 5);
Repeatability (%)

significant differences were found when comparing

respiration and COD removal rates of sludge with

4.0
0.9

0.9

5.7
0.8
26.7
27.8
32.1
25.6
22.9
31.3
}

and without nitrogen/phosphorus balancing, and

comparing normalised data obtained with real and
synthetic greywater. The results indicate that the
Oxygen uptake

origin of the greywater (i.e. real or synthetic) has an

1
d )

effect on the normalised data (i.e. the extent of

1

variation between a test sample and its concurrent

Rate (kg kg MLSS

control), although there was no significant effect on

1.1
1.1
0.9
1.2
1.3
0.9
0.5
0.4
0.6
0.6
0.7
0.6
1.0
0.9

the rates per unit MLSS per day. The t-test results
also show that the presence or absence of nitrogen/
phosphorus balancing in the respirometer does not
affect the normalised data, but has a significant effect
on the rates of uptake per unit MLSS.
Nutrient dose

The t-test results show that the origin of the

Mo
Co
Cu

greywater affects the behaviour of the sludge in

Zn
Fe

Al

response to nutrient dosing, in terms of the magni-

P,
P,
P,
P,
P,
P,

Mo
Co
Cu

Zn
Fe
N,
N,
N,
N,
N,
N,

Al

tude of the differences between test sludge samples

P
2706 Bruce Jefferson et al.

Table 5. Paired t-tests performed on respiration and COD removal data

Oxygen-uptake rate COD-removal rate Normalised oxygen Normalised COD

uptake removal

Real vs. synthetic greywater df=14 df=14 df=13 df=13

t stat=
5.47 t stat=
8.23 t stat=2.27 t stat=3.40
t crit=1.08 t crit=1.08 t crit=1.08 t crit=1.08
p=0.95 p=0.95 p=0.95 p=0.95
No significant No significant Significant difference Significant difference
difference difference
N/P-balanced vs. N/P-limited df=11 df=11 df=11 df=11
greywater
t stat=7.34 t stat=4.54 t stat=
0.02 t stat=
0.03
t crit=1.09 t crit=1.09 t crit=1.09 t crit=1.09
p=0.95 p=0.95 p=0.95 p=0.95
Significant difference Significant difference No significant difference No significant difference

Table 6. Summary of effects of nutrient addition

Nutrient Real greywater Synthetic greywater

With N/P balancing Without N/P balancing With N/P balancing Without N/P balancing

Co Stimulatory Stimulatory Chemical Chemical

Cu Stimulatory Stimulatory Inhibitory Stimulatory
Fe Chemical Stimulatory Stimulatory Chemical
Mo Stimulatory Stimulatory Stimulatory Chemical
Zn Stimulatory Chemical Inhibitory Stimulatory
Al Inhibitory Chemical Inhibitory Stimulatory
P } Stimulatory } Chemical
N } Stimulatory } Stimulatory

Fig. 1. Real greywater: normalised mean oxygen uptake and COD-removal rates.

and their concurrent controls. In this case, biomass
receiving synthetic greywater showed a greater
response to nutrient dosing than sludge supplied
with real greywater. It can also be seen that the
addition of nitrogen/phosphorus dosing does not
affect the increase in COD removal and respiration
rates in test cells compared to their controls, but that
nitrogen/phosphorus-balancing results in signifi-
cantly different COD removal and respiration rates
per unit MLSS regardless of the source of the
greywater.
REPRODUCIBILITY OF TRENDS
The measurements of MLSS, oxygen utilisation
and influent and effluent COD made were used to
calculate respiration and COD-removal rates per unit
 Nutrient addition in greywater
Fig. 2. Synthetic greywater: normalised mean oxygen uptake and COD-removal rates.
MLSS. The repeatability of the duplicates was
calculated by dividing the difference between dupli-
cates by twice the mean, a method which indicates
the percentage repeatability of a measurement.
Percentage values in the range
10.0 to +10.0%
were taken as representing good data reproducibility.
The rates calculated for oxygen uptake and COD
removal per unit MLSS per unit time showed that
sludge supplied with synthetic greywater required
more oxygen and degraded more COD than the
sludge receiving real greywater, under every dosing
regime (Table 4), although a t-test showed this
difference not to be significant (Table 5). It can also
be seen that both COD removal and oxygen uptake
rates increase on the addition of nitrogen/phos-
phorus to any dosing regime, indicating significantly
increased metabolic activity of the sludge. The results
obtained during tests using unbalanced greywater
were more reproducible than tests using nitrogen/
phosphorus-balanced greywater, and synthetic re-
sults were more reproducible than real results (Table
4). Results from tests using cobalt, iron, molybdenum
and aluminium varied most with both types of
greywater. In all cases except zinc addition without
nitrogen/phosphorus balancing, oxygen-uptake rates
were found to be less reproducible than COD-
removal rates. The most striking trend was the lack
of reproducibility in data obtained from nitrogen/
phosphorus-balanced tests compared to data from
unbalanced greywater tests.
DISCUSSION
Effects of micronutrient additions
There is a paucity of published information
regarding the mechanistic roles of micronutrients in
aerobic bacterial cells, while chemical effects of metal
ions on activated sludge in other wastewater applica-
2707
tions are well known. The addition of nitrogen to real
and synthetic greywaters, and of phosphorus to real
greywater resulted in stimulation of the sludge
biomass, as expected when biomass operates under
nutrient-limited conditions. However, adding phos-
phorus to synthetic greywater led to effects indicative
of adsorption/coagulation mechanisms. This suggests
that sludge receiving synthetic greywater was under
nitrogen-limited conditions, and whilst also deficient
in phosphorus the overall imbalance remains nitro-
gen controlled. While the water quality determinands
of synthetic greywater are a good match for real
greywater, the two substrates are nutritionally
different. Real greywater is more variable than
synthetic greywater, as the occasional use of a large
number of different hygiene products affects real
greywater composition. The C : N : P ratio of real
greywater will be affected to a certain extent by the
use of detergents with and without phosphate, but as
such detergents contribute only a part of the
contaminant load, and flow balancing is an inevitable
feature of a greywater treatment collection and
treatment system, they are not expected to be
dominant factor in the characteristics of real grey-
water.
All of the trace metals added to greywater were
found in concentrations lower than the theoretical
requirements (Table 1), and so stimulatory effects
could be expected from their addition. Cobalt,
copper and molybdenum stimulated the bacteria in
real greywater, whether nitrogen/phosphorus ba-
lanced or not. However, while molybdenum led to
stimulation in nitrogen/phosphorus-balanced syn-
thetic greywater, and copper was stimulatory in
unbalanced synthetic greywater, inhibitory effects
were seen in other cases of molybdenum, copper and
cobalt addition. The reasons for these effects are not
certain; cobalt is a metallic enzyme activator required
by some bacteria which synthesise cyanocobalamin
2708 Bruce Jefferson et al.
(vitamin B12) and is also used in the production of
carboxypeptidase (Wood and Tchobanoglous, 1975).
Cobalt chloride and cobalt sulphate can stimulate the
synthesis of cyanocobalamin and some other B12-like
factors in aerobically fermenting sludge (Sathyanar-
ayana Rao and Srinath, 1961), suggesting a stimula-
tory action on metabolism. Doses of cobalt have
shown potential in optimising the activated sludge
process with the addition of 51.0 mg l
1 cobalt.
However, increases in metabolic activity may not be
associated with any improvement in BOD or COD
removal, nor linked to any change in the nitrogen
and phosphorus content of the sludge (Sathyanar-
ayana Rao and Srinath, 1961). These observations on
nitrogen and phosphorus interactions are corrobo-
rated in real greywater but contradicted in tests with
synthetic greywater. Cobalt has also been reported to
be toxic to activated sludge (Geradi, 1986; Hunter
et al., 1966); such inhibitory effects were not noted in
this study.
Copper is found in wastewater streams from
industrial and domestic sources and exerts a toxic
effect on activated sludge (Gracia et al., 1994), an
effect noted when copper was added to nitrogen/
phosphorus-balanced synthetic greywater. However,
copper addition in real and nitrogen/phosphorus-
limited synthetic greywater resulted in metabolic
stimulation, as expected in copper-limited substrates.
Dosing low concentrations of copper into wastewater
has been found to be beneficial in certain cases. Work
in which 0.6–3.0 mg l
1 copper was added to indus-
trial wastewater showed that concentrations above
the accepted toxicity threshold can chelate other
toxicants in the effluent and improve process
performance (Vandevivere et al., 1997). However,
the effects of copper addition fade as the metal
becomes adsorbed onto other binding sites (Vande-
vivere et al., 1997), so it is possible that the
stimulatory effects noted here were temporary.
Molybdenum can commonly be a limiting nutrient
(Grau, 1991), and this appears to be the case in
greywater. In this study, addition of molybdenum
had beneficial effects on sludge supplied with all
greywaters except unbalanced synthetic greywater,
where no conclusive effect was seen. Molybdenum
has been found to have stimulatory effects on
activated sludge treating recalcitrant industrial waste-
water (Burgess et al., 1999b) although little is known
of its biochemical role.
The mechanism of the action of aluminium in
aerobic bacteria is not known, and it appears from
these results that if aluminium is required, the
toxicity threshold is lower than previously thought.
Recently published work based on aluminium dosing
has indicated beneficial effects on activated sludge
treatment of domestic wastewater; BOD and COD
removal was improved (Clark et al., 1999a). Clark’s
results contradict those in the current study, as COD
removal was decreased on the addition of aluminium,
but they refer to coagulant concentrations of 2–3
orders of magnitude above those used in the current
study, where chemical rather than biological mechan-
isms predominate. The inhibitory effect on respira-
tion and COD degradation was seen in all cases but
unbalanced real greywater.
Small stimulatory effects were observed on the
addition of iron to nitrogen/phosphorus-balanced
synthetic greywater. In all other cases, adsorption
appeared to be the dominant process in the iron-
dosed sludge. Iron is an important element in
activated sludge in small quantities. Fe3+ reducing
bacteria are found in most activated sludge mixed
cultures and Fe3+ and sulphate reduction are
important in floc formation. Iron-based co-precipi-
tants are commonly used in aerobic biological
wastewater treatment, as Fe3+ forms strong com-
plexes with the condensed phosphates, pyropho-
sphate and tripolyphosphate, which may then be
removed by adsorption onto Fe3+ hydroxo-phos-
phate surfaces (Jenkins et al., 1971). Work in which
iron-based co-precipitants were added to an activated
sludge system found adverse effects on COD and SS
removal, although the most notable changes were to
floc size and structure rather than effects on the
microbial population (Clark et al., 1999b). The
current study, in which the effects of iron indicate
chemical reactions rather than biological effects,
corroborates previous literature and suggests that
iron may not be a limiting nutrient in terms of
bacterial metabolism.
Zinc has been demonstrated to non-competitively
inhibit bacterial metabolism (Gökcay and Yetis,
1996) and in these instances inhibited enzyme kinetics
can be applied to the metabolism of the cells. Zinc
interacts with other metals (e.g. copper) to exacerbate
their effects and has been shown to reduce the rates of
reactions so that the kinetics of biodegradation in the
presence of zinc and copper resembled those of
biodegradation without the metals but with a very
short sludge age (Beyenal et al., 1997). Clearly the
action of zinc on sludge biomass is strongly
influenced by the chemical conditions and as zinc
and copper were detected in real, but not in synthetic
greywater during ICP analysis it is reasonable to
expect the interactions occurring in the two types of
greywater to differ.
Reproducibility of data
The measurements taken during tests with real
greywater showed more variability than those from
synthetic greywater. The difference in reproducibility
can be accounted for by the nature of the greywater.
The real greywater was collected daily, but its
composition was affected by trends in water use in
the source building, the wider range of possible
constituents and the water level in the collection
sump at the time of sampling. The source used here
was naturally low in organics (mean BOD 48 mg l
1)
which made the absolute rates of degradation and
 Nutrient addition in greywater
respiration lower than with the synthetic greywater.
An additional factor was the length of time a sample
of real greywater might spend in the reservoir from
which the daily samples were taken, as previous work
has shown real greywater to degrade rapidly (Jeffer-
son et al., 1999). Real greywater is highly site specific
in nature, with large variations in standard constitu-
ents such as BOD and COD. Greywater sampled
from a range of sites around the UK has shown BOD
variations from 30 to 300 mg l
1 (Jefferson et al.,
1999). Synthetic greywater acts as a good analogue of
this. However, the results presented here show
significant differences between the sources of
greywater at the nutrient level. Thus, although
general trends appear consistent, complex nutrient
addition will undoubtedly be site specific. These
results imply that using nutrient addition in biologi-
cal treatment of greywater may not produce results as
consistent as those obtained in studies that employ
synthetic greywater to predict treatability in grey-
water recycling systems. More variability was ob-
served in the measurements of oxygen-uptake rate
than in COD-removal rate with both real and
synthetic greywaters.
Less repeatable results were obtained for cobalt,
iron, molybdenum and aluminium than for other
micronutrients. This may be owing to their action on
biomass. Iron and aluminium are known to have
biochemical roles as well as varying coagulation
effects that are dependent on the prevailing chemical
conditions. All four metals exhibited effects ranging
from stimulatory to inhibitory in the current study,
and have inconsistent effects on other biological
wastewater treatment processes (e.g. cobalt–stimula-
tory effects, Wood and Tchobanoglous, 1975;
Sathyanarayana Rao and Srinath, 1961}inhibitory
effects, Geradi, 1986; Hunter et al., 1966). A decrease
in reproducibility was observed when greywaters
were nitrogen/phosphorus balanced and then dosed
with micronutrients. This may be because the
unbalanced greywaters were both nitrogen/phos-
phorus-limited, such that nitrogen/phosphorus-bal-
ancing caused another characteristic of the greywater
to become the limiting factor.
Impacts on process performance
Nitrogen/phosphorus balancing made a significant
difference to the rates of oxygen uptake and COD
removal in both real and synthetic greywaters,
indicating that nitrogen and phosphorus play vital
roles in the biodegradation of greywater pollutants.
Increased COD removal rates were observed in 24 of
the 28 nutrient dosing regimes tested, suggesting that
macro- and micro-nutrient balancing will have a
potential role in greywater treatment and recycling in
the future. The effects of dosing with additional
nutrients, beyond simple nitrogen and phosphorus
balancing, are unclear. The relative normalised
effects of each of the complex nutrients were
2709
independent of nitrogen and phosphorus balancing.
Further, variations in results increased once nutrient
balancing had occurred. This indicates that either
* after nitrogen and phosphorus balancing, other
substances than those tested were imposing
process limitations,
or
* potentially toxic overdoses due to undetected
variations in the existing background concentra-
tions of micro-nutrients that were being dosed.
Although the source of the greywater did not
significantly affect the rates of COD removal or
oxygen uptake, it did affect the relative normalised
difference between real and synthetic greywaters.
This highlights the site-specific nature of the nutri-
tional characteristics of greywater, thus revealing a
limitation in the use of synthetic grey as a good
medium for complex treatability studies.
Overall, significant improvements in greywater-
treatment process efficacy are probable with simple
nitrogen/phosphorus-balancing. Further, improve-
ments are possible with more complex nutrient
dosing but these benefits appear more spurious.
Although the independence of the effects of the
complex nutrients to nitrogen/phosphorus balancing
indicate definite effects, the much greater variation
in the results after nitrogen/phosphorus balancing
decrease the confidence with which these results
could be exploited.
The implications for greywater treatment are that
nitrogen/phosphorus balancing alone allows the
greatest potential for process enhancement. This is
likely to manifest itself through increasing the
maximum loading rates that can be applied without
compromising effluent quality.
CONCLUSIONS
Nutrient balancing has the potential to improve
biological treatment of greywater, as nitrogen/
phosphorus-balancing eliminates the occurrence of
macronutrient-limiting conditions that often impede
substrate degradation. The data from the current
study demonstrate that biological treatment of the
real greywater employed can be maximised by
nitrogen/phosphorus balancing followed by supple-
ments of zinc or copper. Results from this study led
to the conclusions that
* Nitrogen/phosphorus balancing can improve bio-
logical treatment of greywater.
* Micronutrient addition can improve biological
greywater treatment, but additions must be
tailored to the greywater in each case.
* Synthetic greywater does not react to nutrient
addition in the same way as real greywater.
* Complex interactions are evident during micro-
nutrient dosing.
2710 Bruce Jefferson et al.
* The effects of micronutrients are not confined to
the biology of the treatment system alone.
Acknowledgements}The authors would like to express their
gratitude to EPSRC, Thames Water Utilities and Yorkshire
Water Services Ltd. for financial support of this research.
REFERENCES
Annaka T. (1977) Upgrading of existing sewage treatment
plants by chemical addition to aeration tank. Proceedings
of the fifth US-Japan Conference on Sewage Treatment
Technology, Tokyo, April, pp. 1–17.
APHA (1992) Standard Methods for the Examination of
Water and Wastewater, 18th ed. American Public Health
Association, Washington, DC, USA.
Beardsley M. L. and Coffey J. M. (1985). Bioaugmentation:
optimizing biological wastewater treatment. Pollut.
Engng. December, 30–33.
Benefield L., Lawrence D. and Randall C. (1979) The effect
of sludge viability on biokinetic coefficient evaluation.
J. Water Pollut. Control Fed. 51, 187–194.
Beyenal N. Y., Özbelge T. A. and Özbelge H. O. (1997)
Combined effects of Cu2+ and Zn2+ on activated sludge
process. Water Res. 31, 699–704.
Blackall L. L., Seviour E. M., Bradford D., Stratton H. M.,
Cunningham M. A., Hugenholts P. and Seviour R. J.
(1996) Towards understanding the taxonomy of some of
the filamentous bacteria causing bulking and foaming in
activated sludge plants. Water Sci. Technol. 34(5–6),
137–144.
Burgess J. E., Quarmby J. and Stephenson T. (1999a) The
role of micronutrients in biological treatment of industrial
effluents using the activated sludge process. Biotechnol.
Adv. 17, 49–70.
Burgess J. E., Quarmby J. and Stephenson T. (1999b)
Micronutrient supplements to enhance biological waste-
water treatment of phosphorus-limited industrial efflu-
ents. Trans. Inst. Chem. Eng. 77(B), 199–204.
Clark T., Stephenson T. and Arnold-Smith A. K. (1999a)
The impact of aluminium based co-precipitants on the
activated sludge process. Trans. Inst. Chem. Eng. 77(B),
31–36.
Clark T., Burgess J. E., Stephenson T., and Arnold-Smith
A. K. (1999b) The influence of iron-based co-precipitants
on activated sludge biomass. Trans. Inst. Chem. Eng.
78 (B5), 405–410.
Franta J., Wilderer P. A., Miksch K. and Sykora V. (1994)
Effects of operation conditions on advanced COD
removal in activated sludge systems. Water Sci. Techol.
29(7), 189–192.
Geradi S. M. H. (1986) Effects of heavy metals upon the
biological wastewater treatment process. Public Works
117(6), 77–80.
Gökcay C. F. and Yetis U. (1996) Effect of nickel(II) on the
biomass yield of the activated sludge. Water Sci. Technol.
34(5–6), 163–172.
Gostick N. (1991) A study of the effect of substrate
composition on the settlement of activated sludge.
Ph.D. Thesis, Cranfield University, UK.
Gracia M. -P., Salvadó H., Ruis M. and Amigo J. M. (1994)
Effects of copper on ciliate communities from activated
sludge plants. Acta Protozool. 33(4), 219–226.
Grau P. (1991) Criteria for nutrient - balanced operation of
activated sludge process. Water Sci. Tech. 24(3–4),
251–258.
Heddle J. F. (1980) Respirometric oxygen demand tests for
wastewater. Proceedings of the Aquatic Oxygen Seminar,
Hamilton, November. Published by the Water and Soil
Division, Ministry of Works and Development, Well-
ington, New Zealand. Water and Soil Miscellaneous
Publications, no. 29.
Hunter J. V., Genetelli E. J. and Gilwood M. E. (1966)
Temperature and retention time relationships in the
activated sludge process. Proceedings of the 21st Industrial
Waste Conference, Purdue University, Indiana, vol. 121,
pp. 953–963.
Jefferson B. and Judd S. (1999). Water recycling: asking the
right questions? Water Environ. Int. 15B, 15–16.
Jefferson B., Laine A. and Judd S. (1999) Technological
requirements for water recycling. Environ. Prot. Bull. 60,
8–14.
Jenkins O., Fergusson J. F. and Menar A. B. (1971)
Chemical processes for phosphate removal. Water Res. 5,
369–387.
Laine A., Jefferson B., Judd S. and Stephenson T. (1999)
Membrane bioreactors and their role in wastewater reuse.
Water Sci. Technol. 41(1), 197–204.
Metcalf and Eddy, Inc. (1991) Wastewater Engineering}
Treatment, Disposal and Reuse. In McGraw-Hill Series in
Water Resources and Environmental Engineering, 3rd ed.
G. Tchobanoglous and F. L. Burton. (Eds.), Macgraw
Hill, New York.
Nicholas D. J. D. (1963). Inorganic nutrient nutrition of
microorganisms. In: Plant Physiology III, ed. F. C
Stewart, Academic Press, New York, pp. 363–375.
Pala A. I. and Sponza D. T. (1996) Biological treatment of
petrochemical waste waters by Pseudomonas spp. added
activated sludge culture. Environ Technol. 17, 673–686.
Sathyanarayana Rao S. and Srinath E. G. (1961) Influence
of cobalt on the synthesis of vitamin B12 in sewage during
aerobic and anaerobic treatment. J. Sci. Ind. Res. 20(c),
261–265.
Singleton I. (1994) Microbial metabolism of xenobiotics:
fundamental and applied research. J. Chem. Technol.
Biotechnol. 59(1), 9–23.
Soddell J. A. and Seviour R. J. (1996) Growth of an
activated sludge foam-forming bacterium, Nocardia
pinensis on hydrophobic substrates. Water Sci. Technol.
34(5–6), 113–118.
Speitel G. E. and Digiano F. A. (1988) Determination of
microbial kinetic coefficients through measurement of
initial rates by radiochemical techniques. Water Res. 22,
829–835.
Speitel G. E. and Segar R. L. (1995) Cometabolism in
biofilm reactors. Water Sci. Technol. 31(1), 215–225.
USEPA (1996) Prevention, Pesticides and Toxic Substances
(7101). EPA 712-C-96-301. Fate, transport and trans-
formation test guidelines. OPPTS 835.3220 (Porous pot
test). USEPA, Cincinnati.
Valo R., Apajahalati J. and Salkinoja-Salonen M. (1985)
Studies on the physiology of microbial degradation of
pentachlorophenol. Appl. Microbiol. Biotechnol. 21,
313–319.
Vandevivere P., Ficara E., Julies E. and Verstraete W.
(1997) Adding copper to detoxify wastewater. Proceedings
of the Environmental Technology International Symposium
April 21–23, Gent, Belgium, 405–411.
Water Research Centre (1978) Porous Pot Method for
Assessing Biodegradability. Technical Report TR70,
Water Research Centre, Medmenham, UK.
Wood D. K. and Tchobanoglous G. (1975) Trace elements
in biological waste treatment. J. Water Pollut. Control
Fed. 47, 1933–1945.