Nitrogen Removal

Prof. Dr. George A. Ekama

University of Cape Town
SOUTH AFRICA

Dr. Mark C. Wentzel

University of Cape Town
SOUTH AFRICA

Sponsored by

1
 BIOLOGICAL
NITRIFICATION

George Ekama
Mark Wentzel
Water Research Group
Dept of Civil Engineering
University of Cape Town

OUTLINE (1)
ƒ GENERAL
ƒ BIOLOGICAL KINETICS
ƒ Growth behaviour
ƒ Endogenous respiration
ƒ PROCESS KINETICS
ƒ Effluent ammonia concentration
ƒ Minimum sludge age for nitrification

2
OUTLINE (2)
ƒ FACTORS INFLUENCING
NITRIFICATION
ƒ Wastewater source
ƒ Temperature
ƒ Unaerated zones
ƒ Dissolved oxygen concentration
ƒ Cyclic flow and load
ƒ pH and alkalinity
ƒ NUTRIENT REQUIREMENTS FOR
SLUDGE PRODUCTION

OUTLINE (3)
ƒ DESIGN CONSIDERATIONS
ƒ Effluent TKN concentration
ƒ Nitrification capacity
ƒ NITRIFICATION DESIGN EXAMPLE
ƒ Effect of nitrification on reactor pH
ƒ Minimum sludge age for nitrification
ƒ W t
Wastewater t N concentrations
t ti
ƒ Nitrification process behaviour.

3
NITRIFICATION (1)
ƒ The process – biological oxidation of free
(NH3) and saline (NH4+) ammonia (FSA)
to nitrite (NO2-) and nitrate (NO3-) in two
sequential steps by two groups of
obligate aerobic autotrophic organisms.
ƒ NH4+ + 3/2O2 → NO2- + H2O + 2H+ + Energy
Ammonia oxidizing organisms - AOOs
ƒ NO2- + 1/2O2 → NO3- + Energy
Nitrite oxidizing organisms - NOOs

NITRIFICATION (2)
ƒ Energy is used in catabolism for new cell
mass synthesis (anabolism) -
ƒ Energy + NH4+ + CO2 + H2O → Biomass
ƒ Yield of biomass for autotrophs very low
YA~0.10 mgVSS per mgFSA
mgFSA--N nitrified
ƒ Also NOOs (Step 2) much faster than
AOOs (Step 1) - no NO2- build up.
ƒ So can model nitrification as single
process (Step 1 + Step 2) at Step 1 rate.

4
NITRIFICATION (3)
Overall process stoichiometry by autotrophic
nitrifier organisms (ANOs)
NH4+ + 2O2 → NO3- + H2O + 2H+

From above – per mgFSA

mgFSA--N nitrified -
ƒ 4.57 mgO utilized (64mgO/14mgN)
ƒ 7.14 mg/l as CaCO3 consumed (2x50/14)
ƒ 1 mgNO3--N generated.

NITRIFICATION (4)
If ammonia uptake by ANOs for biomass
growth is taken into account, then
products
d t are negligibly
li ibl llower -

Per mgFSA-
mgFSA-N nitrified -
ƒ 4.45 gO utilized (4.57mgO/mgN)
ƒ 7.11 mg/l as CaCO3 consumed (7.14)
ƒ 0.99 mgNO3--N generated (1.00).

5
NITRIFIER GROWTH KINETICS (1)
Based on Monod growth kinetics –
(1) ANO biomass (XBA) generated is a fixed
fraction (YA) of FSA (Na) nitrified –
dXBA dN a
' YA &
dt dt
(2) Specific growth rate (μ
(μAT) related to bulk
liquid FSA concentration (Na) - Monod
dXBA :AmT Na
' :AT XBA ' XBA
dt KnT %Na

NITRIFIER GROWTH KINETICS (2)

μAm = 0.45/d
0 45/d
Kn =1 mgN/l
YA = 0.10
mgVSS/
mgFSA--N
mgFSA

If Na>4, ANOs nitrifying at maximum rate (μ (μA),

but it’s difficult to get low Na<1.0 mgN/l

6
NITRIFIER GROWTH KINETICS (3)

FSA (Na) utilization rate, NO3- (Nn)

generation rate and nitrification oxygen
g yg
utilization (On) rate are linked to ANO
biomass (XBA) growth rate -
dNn
' &
dNa
' 1 : Am T N a X
BA
dt dt Y A K nT % N a

dOn dNa dNn

' 4.57 ' 4.57
dt dt dt

NITRIFIER ENDOGENOUS
RESPIRATION
Like all organisms, ANOs also undergo a biomass
loss due to maintenance or endogenous energy
requirements.
This is modelled in the same way in both steady
state and simulation models as endogenous
respiration for the ordinary heterotrophic
organisms (OHOs) in steady state model, viz.
bAT = specific
dXBA endogenous mass
' &b AT X BA loss rate at ToC
dt = 0.04 /d at 20oC

7
PROCESS KINETICS
Derived from mass
balances over AS system
(1) Mass balance on XBA -
)X BA :AmT Na QW
Vp ' Vp X BA & Vp bAT X BA & Vp X BA ' 0
)t KnT %N a Vp
at steady
(2) Solving for Na (= Nae) yields - state
KnT(bAT % 1/ RS) Independent
Na ' Nae '
: AmT & (bAT % 1/RS) of YA and Nai

MINIMUM SLUDGE AGE (1)

Plotting Nae vs Rs
(1)Above
(1) Above certain Rs,
Nae is veryy low.
(2) As Rs decreases,
Nae increases.
(3) When Nae = Nai,
Rs ≈ Rsm.
Rsm = minimum sludge age for nitrification.
Sludge age (Rs) is most important design
parameter for systems required to nitrify!

8
 MINIMUM SLUDGE AGE (2)
(1) If Rs < Rsm, no nitrification.
(1)If
(2)If
(2) If Rs > 1
1.3
3 x Rsm, nitrification almost
complete.
Setting Rs = Rsm and Nae=Nai yields for Rsm
1
RSm ' and with 1
K RSm '
(1% nT ) :AmT & bAT KnT << Nai :AmT & bAT
Nai

So Rsm depends mainly on maximum

specific growth rate of ANOs - µAm

NITRIFIER µAm
Depends on many factors -
((1)) Wastewater
aste ate composition
co pos t o (metals,
( eta s, salts).
sa ts)
(2) Wastewater temperature (T↓, μAm↓).
(3) Wastewater pH (pH↓, μAm↓).
(4) Reactor DO concentration (DO↓, μAm↓).
(5)ANO
(5) ANO population selection
selection.
Therefore µAm is considered a wastewater
characteristic rather than kinetic constant

9
FACTOR OF SAFETY ON µAm
Nitrification is prerequisite for N removal, so
to ensure nitrification, µAm is decreased
by a factor of safety
safety, Sf (1.2
(1 2 – 1.3).
1 3) This
(1) ensures Rs > Rsm,
(2) covers uncertainty in µAm
(3) ensures low effluent FSA concentration
andd near complete
l t nitrification.
it ifi ti
and under constant flow and load (steady
state) Nae = KnT / (Sf - 1)
1)..

FACTORS AFFECTING
NITRIFICATION
(1) wastewater – magnitude of µAm20,
(2) temperature,
(3) unaerated zones in reactor,
(4) aerobic reactor DO concentration,
(5) cyclic flow and load conditions
(6) reactor pH.

Each is discussed further below.

10
(A) WASTEWATER SOURCE
ƒ Already mentioned - µAm20 varies much
between different wastewaters and so is a
wastewater characteristic rather than a
kinetic constant.
ƒ Ideally should be measured on wastewater.
ƒ µAm20 values range
g between 0.3 - 0.75 /d.
ƒ bA20 is accepted to stay constant (0.04/d).

(B) TEMPERATURE (1)

Nitrifier kinetic constants µAm20, bA20 and Kn20
all decrease with temperature
temperature.
(1) µAmT = µAm20(Θn)(T (T--20) ; Θ =1.123
n
(2) KnT = Kn20(Θn) (T-
(T -20) ; Θn=1.123
(3) bnT = bn20(Θb)(T(T--20) ; Θ =1.029.
b

Θn=1.123 is equivalent to a 50%

reduction every 6oC – if 0.45/d at
20oC, then is 0.23 at 14oC.

11
(B) TEMPERATURE (2)
ƒ Halving µAm by 6oC decrease, doubles
minimum sludge age for nitrification (Rsm).
)
ƒ Halving Kn by 6oC decrease does not affect
Rsm, but does affect effluent FSA conc.
ƒ Overall effect of µAm and Kn decrease with
temperature is effluent FSA increase – the
lower the temp, the higher the FSA conc.

(C) UNAERATED ZONES (1)

ƒ Formulated on 3 assumptions -
((1)) ANOs g grow onlyy in aerobic zone
(2) ANO endogenous respiration occurs in
all zones at same rate (controversial),
(3) Proportion of ANOs in VSS same in all
zones..
zones
ƒ Important implication of (3) is that sludge
mass fractions of different zones reflect
distribution of ANOs in system.

12
 (C) UNAERATED ZONES (2)
ƒ From 3 assumptions it can be shown that -
KnT ((bAT % 1/Rs)
Nae '
µAmT (1&fxt) & (bAT % 1/Rs)
fxt = unaerated mass fraction,
= fraction of total mass of sludge in
system
y in unaerated zone(s).
( )
= 0 for fully aerobic system.
ƒ Similarly, Rsm ' 1
µ AmT (1&fxt) & bAT

(C) UNAERATED ZONES (3)

Alternatively, the maximum unaerated
sludge mass fraction fxm allowed at a
sludge age of Rs to ensure nitrification
with a safety factor of Sf is –
fxm = 1 - Sf (bAT + 1/Rs)/µAmT

and the effluent FSA

conc is still given by
Nae = KnT / (Sf - 1)

13
 (D) DO CONCENTRATION
ƒ Effect of DO on µAm is formulated as -
O
µAmO ' µAm
KO % O
O = DO conc in mixed liquor
KO = Monod half saturation conc for DO
= 0.3 to 2 mgO/L (depends on floc
size, mixing).
ƒ If reactor DO < KO, nitrification rate is
less than half the maximum.

(E) CYCLIC FLOW AND LOAD

(1) Average effluent FSA conc under cyclic flow
and load > steady state effluent FSA conc.
(2) So nitrification efficiency decreases under
cyclic flow and load compared with steady
state - the higher the variation, the lower the
efficiency.
((3)) Cyclic
y flow and load has similar effect as
decreasing sludge age closer to Rsm
(4) So iincrease
ncrease Rs to compensate for cyclic
flow and load and keep effluent FSA low.

14
(F) pH and ALKALINITY (1)

ƒ Nitrification
consumes 7 7.14
14
mg/l Alk as CaCO3
per mgN/l FSA
nitrified.

ƒ If mixed liquor alkalinity decreases

below 40 mg/l as CaCO3, mixed
liquor pH decreases below 7.

(F) pH and ALKALINITY (2)

ƒ Nitrification is
very sensitive to
pH.
ƒ Optimum pH
range is 7-7-8.
ƒ In low alkalinity WW
WW, nitrification can inhibit
itself due to H+ release, which reduces mixed
liquor pH below 7, which reduces μAm.

15
 (F) pH and ALKALINITY (3)

ƒ If influent Alk = 200 and 24 mgN/l FSA

is nitrified
nitrified, effluent Alk = 200 –
7.14x24=29 mg/l. Less than 40, so
mixed liquor pH will decrease below 7.
ƒ In this event, either
ƒ Introduce anoxic zones to denitrify
y nitrate
and recover half Alk lost, or
ƒ Dose lime to keep pH > 7.0.

N REQUIREMENTS FOR
SLUDGE GROWTH
(1)About
(1) About 15-
15-20% of
influent TKN is
required for AS Nte
sludge growth (N(Ns).
(2) Ns decreases with NC NS
Rs and settled WW.
(3)Influent
(3) Influent biodeg OrgN
adds to FSA pool in Nti
reactor and nitrified.

16
 FATE OF INFLUENT TKN
Released as Escapes
FSA and with
adds
dd tto effluent
FSA pool.
Some taken up
for growth,
rest nitrified
to nitrate.
Effluent FSA Effluent TKN Included with
(Nae) low. Nte=Nae+Nousi waste sludge

NITRIFICATION CAPACITY
Nitrification capacity (Nc) = Concentration of
nitrate per litre influent flow generated by
nitrification (mgN/l).
Nc = Nti – Ns – Nte ; Ns = fn (Vp.Xv)/(Rs.Qi)
Nte= Nae + Nousi ; fn = N content of VSS
KnT (bAT % 1/Rs)
Nae '
µAmT (1&fxt) & (bAT % 1/Rs)

17
DESIGN PROCEDURE (1)
Calculation of sludge age (Rs) and
unaerated sludge mass fraction (fxt) most
important part of design procedure!
With WW characteristics known and Rs and
fxt selected -
(1) Calculate influent N concentrations
(Noi, Nousi, Noupi).
)
(2) Calculate effluent OrgN (Nouse = Nousi).
(3) Calculate N for sludge production (Ns).

DESIGN PROCEDURE (2)

(4) Calculate Rsm.
(5) If Rs < Rsm – no nitrification, then -
Effluent Nitrate (Nne) = 0.
Effluent FSA (Nae) = Nti – Ns - Nouse
= FSA available for nitrification (Nan).
Effluent TKN (Nte) = Nae + Nouse = Nti - Ns
Nitrifier sludge mass MXBA = 0
Nitrification oxygen demand MOn = 0

18
DESIGN PROCEDURE (3)
(6) If Rs > Rsm – nitrification occurs, then –

KnT (bAT % 1/Rs)

Effluent FSA = Nae '
µAmT (1&fxt) & (bAT % 1/Rs)

If Nae is negative or >Nan, set Nae = Nan

Effluent TKN: Ntet = Nae + Nouse (mgN/l)
Effluent Nitrate: Nne = Nan – Nae (mgN/l)
= Nti – Ns – Nte (mgN/l)

DESIGN PROCEDURE (4)

(6) If Rs > Rsm – nitrification occurs, then –
Analogous to mass of ordinary heterotrophic
organisms (OHOs) in reactor, mass of
ANOs is MXBA = (Qi Nne) YARs/(1+bATRs)
where Qi Nne = mass of nitrate generated
= Nne (Qe + Qw) mgN/d
Nitrification oxygen demand (mgO/d) =
MOn = 4.57 (Qi Nne) mgO/d.

19
EXAMPLE - RAW WW
mgN/l 60*
Nousi=Effl TKN-
TKN-FSA

15.0 0.45μ filtered - Sol

0.45μ
TKN =48.5 mgN/l
45* 1.7 1.8** 48.5*
fN’ous= Nousi/Nti= 0.03
Nousi=0.03x60

39
3.9 7.6 =1.8 mgN/l
11.5 Noupi= fnSupi/fcv
= 0.1x113/1.48
Measured on *influent and **effluent = 7.6 mgN/l

EX - SETTLED WW
mgN/l 51*
Nousi=Effl TKN-
TKN-FSA

6.0 0.45μ filtered - Sol

0.45μ
TKN =48.5 mgN/l
45* 1.7 1.8** 48.5*
fN’ous= Nousi/Nti= 0.03
Nousi=0.035x51
1.3 1.2 =1.8 mgN/l
2.5 Noupi= fnSupi/fcv
= 0.1x18/1.48
Measured on *influent and **effluent = 1.2 mgN/l

20
DESIGN EXAMPLE
ƒ Following the design procedure with the
example raw and settled wastewaters in
th notes
the t att 14 andd 22oC -

DESIGN RESULTS – 14oC

Fully aerobic – raw and settled wastewater.

Minimum sludge age for nitrification ≈ 5d.
Effluent TKN & NO3 concs virtually the same.
Influent TKN and Ns lower for settled than raw.

21
DESIGN RESULTS – 22oC

Fully aerobic – raw and settled wastewater.

Minimum sludge age for nitrification ≈ 2d.
Effluent TKN & NO3 concs virtually the same.
Influent TKN and Ns lower for settled than raw.

EFFECT OF PRIMARY
SETTLING ON NITRIFICATION
PSTs do not have significant influence on
nitrification because -
(1) Not much N is removed in PSTs (10- (10-
15%) – usually most TKN is FSA,
(2) N for sludge production (Ns) for settled
wastewater is lower than for rawraw.
(1)≈(2) so nitrification capacity (Nc) for
settled and raw wastewaters are similar.

22
NITRIFICATION:
INFLUENCE ON SYSTEM (1)
(1) Sludge age: Nitrification requires Rs>Rsm,
so has major influence on selection of Rs.
(2) Reactor volume and sludge production:
For the same sludge age, no influence.
Nitrifiers < 4% of VSS mass in reactor
(TKN load << COD load and YA<< YH).
However nitrification usually needs
However,
longer Rs so reactor volume larger and
sludge production lower.

NITRIFICATION:
INFLUENCE ON SYSTEM (2)
(3) Oxygen demand (OD): Increases
significantly
i ifi tl withith nitrification
it ifi ti – by
b about
b t
40--60% of COD removal OD depending
40
on influent TKN/COD conc ratio.
Also, if nitrification requires longer sludge
age, COD removal OD increases.
(4) In low DO conditions
conditions,, COD removal OD
takes preference and nitrification will be
partial – DO should be >2 mgO/l.

23
NITRIFICATION:
INFLUENCE ON SYSTEM (3)
(5) Effluent quality: No difference in COD, low
FSA high
FSA, hi h nitrate,
it t reduced
d d alkalinity,
lk li it lower
l
pH – possibly aggressive to concrete
surfaces.
(6) When nitrification can take place, by design
or accident, include denitrification and
hydraulic
y control of sludge
g age,
g especially
p y for
warm WW, to reduce nitrate and oxygen
demand, recover alkalinity, raise pH and
minimize rising sludge in SST.

SUMMARY (1)
ƒ (1) Maximum specific growth rate of nitrifiers
(μAm20) fi
fixes th
the sludge
l d age off ththe AS system.
t
ƒ (2) Unaerated zones, low WW temperature and
cyclic flow and load increase sludge age over
minimum for nitrification.
ƒ (3) Selection of unaerated mass fraction and
sludge age is the most important decision in the
design for nitrification.
ƒ (4) At fixed sludge age, nitrification has
negligible effect on sludge production.

24
SUMMARY (2)
ƒ (5) Nitrification increases oxygen demand by
40 to 60% over that for organic removal
removal.
ƒ (6) When nitrification can take place, include
denitrification to recover half the Alkalinity
and oxygen used in nitrification and minimize
rising sludge in the SST.
ƒ (7) Control sludge age hydraulically by
wasting directly from the reactor to fix sludge
age and “guarantee” nitrification rather than
control reactor concentration (system fails via
high ESS rather than on nitrification).

ACKNOWLEDGEMENTS
University of
Cape Town
And others

25