General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from orbit.dtu.dk on: Dec 20, 2017

Incremental design of control system of SHARON-Anammox process for autotrophicnitrogen removal

Mauricio Iglesias, Miguel; Valverde Pérez, Borja; Sin, Gürkan

Publication date:2012

Document VersionPublisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA):Mauricio Iglesias, M., Valverde Perez, B., & Sin, G. (2012). Incremental design of control system of SHARON-Anammox process for autotrophic nitrogen removal Technical University of Denmark (DTU). [Sound/Visualproduction (digital)]. 10th European Workshop on Advanced Control and Diagnosis, Kgs. Lyngby, Denmark,08/11/2012

Incremental design of control system of SHARON-Anammox process for autotrophic

nitrogen removal

Miguel Mauricio-Iglesias Borja Valverde-Pérez Gürkan Sin

08/11/2012 2

Introduction and motivation

The design of the control structure of a bioreactor is challenging due to:

- Plant and model complexity with different time scales and large interactions

- Lack of actuators/control degrees of freedom to act upon the process

- Narrow operation window (in particular pH, temperature and concentration of substrate/product)

- Model mismatch due to lack of knowledge about the microorganism metabolism

- Adaptation of the microorganisms during operation

We illustrate all these challenges in the design of a control structure for a process of nitrogen removal in wastewater treatment

08/11/2012 3

Overview of the presentation

1) Why autotrophic nitrogen removal?

2) Modelling of the plant

3) Control design. Top-down analysis

4) Control design. Bottom-up design

i. Closed loops disturbance gain

ii. H∞ controller

5) Evaluation

6) Conclusions

08/11/2012 4

Why autotrophic nitrogen removal?

Influent

Anaerobic digester

Carbon addition

Nitrification/denitrification

Autotrophic nitrogen removal

- Less aeration and energy

- Lower footprint and sludge production

- No need of carbon addition

NH4+

NO3-

O2

N2

NH4+

NO2-

O2

N2

NH4+

08/11/2012 5

Model

=

= −∑i

iin out

n

dVF F

dt 1

( ), , ,( )⋅

= ⋅ − ⋅ + ⋅ − ⋅ + ⋅*iin net i in net out i L i i i i

d V C F C F C k a C C V r Vdt

( )++ +⋅

= ⋅ + ⋅ + ⋅ − ⋅in in acid acid base base outd V Z F Z F C F C F Z

dt

( )4 30 += − +TNH NH NH

, 4 4 30 + += ⋅ − ⋅e NHK NH NH H

2 23 3 3 2 4 4 2 40 2 2+ − − − − − − − + += − − − ⋅ − − ⋅ − + +Z NO HCO CO H PO HPO NO OH NH H

Components:

8 compounds

4 microorganism types

charge

Dynamic mass and charge balances. 13 ODEs per reactor

Determination of pH. 11 AEs per reactor

Mass balances (4)

Equilibrium relations (6)

Charge balance

Microbial kinetics

ri 11 processes considered

Vangsgaard et al. 1994

08/11/2012 6

Control design. Top-down analysis

Control objectives

removal of >80 % of N load

Influent

TNH = 700 gN/m3

TNO2/TNH = 1.3

Slow growth ~11 d Faster growth ~1 d

Use the SHARON reactor to dampen the disturbances

Control degrees of freedom (3)

How to link the controlled variables with the control objectives?

08/11/2012 7

Control design. Top-down analysis

Potential controlled variables in SHARON

TNO2

pH

TNO2/TNH DO

TNH

Conflict between regulation and the control objectives

08/11/2012 8

Control design. Bottom-up synthesis

Selection of controlled variables

Closed Loop Disturbance Gain (CLDG)

Hovd and Skogestad 1994

1( ) ( ) ( )s s sdCDLG G G G−=

CLDG = |δi |< 1

The disturbance effect is lower than ∆ymax

CLDG = |δi |>1

The disturbance effect is higher than ∆ymax Need of control action!

0 1 2 3 4 5 6 750

100

150

200

250

300

Time (days)

Flow

rate

(m3 /d

)

0 1 2 3 4 5 6 71445

1450

1455

1460

1465

1470

Amm

oniu

m co

ncen

trat

ion

(g/m

3 )

FlowrateTotal ammonium

0 1 2 3 4 5 6 70

5 x 104

Time (days)

Flow

rate

(m3 /d

)

0 1 2 3 4 5 6 70

50

Amm

oniu

m co

ncen

trat

ion

(g/m

3 )FlowrateTotal ammonium

08/11/2012 9

Control design. Bottom-up synthesis

10-2 10-1 100 101 102 10310-2

100

102

Frequency (rad/day)

Mag

nitu

de

pHTNO2/TNH

10-2 10-1 100 101 102 10310-2

100

102

Frequency (rad/day)

Mag

nitu

de

pH DO

10-2 10-1 100 101 102 10310-2

100

102

Frequency (rad/day)

Mag

nitu

de

pH TNO2

10-2 10-1 100 101 102 10310-2

100

102

Frequency (rad/day)

Mag

nitu

de

pH TNH

Variables

pH DO

pH TNH

pH TNO2

pH TNO2/TNH

DO TNO2/TNH

Variables

TNO2 TNO2/TNH

TNH DO

TNH TNO2

TNO2 DO

TNH TNO2/TNH

08/11/2012 10

Control design. Bottom-up synthesis

Selection of controlled variables

Other requirements

(s) (s)IS=

I+G C

(s) (s)(s) (s)

G CT=I+G C

has to be bounded for performance

has to be bounded for robustness and to avoid sensitivity to noise

(s)(s) (s)CC S=

I+G C has to be bounded to penalize large inputs

Cmin N(C) N

U

T

P

W C S

W T

W S∞

Skogestad, S. and Postlethwaite, I. (2005)

Jahanshahi, E. and Skogestad, S. (2012)

stacked optimal H∞ problem

( )( ) ( )( )1σ CS jω γ σ jω uW −≤

( )( ) ( )( )1σ T jω γ σ jω TW −≤

( )( ) ( )( )1σ S jω γ σ jω PW −≤

γ must be small to ensure good controllability

08/11/2012 11

Control design. Bottom-up synthesis

Selection of controlled variables

Variables γ

pH DO 13.1

pH TNH 48.8

pH TNO2 58.6

pH TNO2/TNH 94.3

DO TNO2/TNH 100

TNO2 TNO2/TNH 100

TNH DO 101

TNH TNO2 101

TNO2 DO inf

TNH TNO2/TNH inf

Ranking

10-2 10-1 100 101 102 10310-2

100

102

Frequency (rad/day)

Mag

nitu

de

pH DO

10-2 10-1 100 101 102 10310-2

100

102

Frequency (rad/day)M

agni

tude

TNO2 DO

10-2 100 102

100

1010

Frequency (rad/day)

Mag

nitu

de

TNHTNO2/TNH

08/11/2012 12

Control design. Bottom-up synthesis

DOC = DO Controller pHC = pH Controller LC = Level Controller

Regulatory layer

TNO2/TNH ratio

N2 conc. (Anammox reactor)

g/m

3 N2

Time (days)

Reponse to +5% step input in inflow

Good regulation

Need of master loop to correct the TNO2/TNH ratio

08/11/2012 13

Control design. Bottom-up synthesis

DOC = DO Controller pHC = pH Controller LC = Level Controller RC = TNO2/TNH Controller

Supervisory layer

TNO2/TNH ratio

N2 conc. (Anammox reactor)

g/m

3 N2

Time (days)

Reponse to +5% step input in inflow

The TNO2/TNH ratio is kept at the optimal value

Need of feedback from Anammox reactor

Model mismatch

Ammonium oxidation due to washed bacteria

0 10 20 30 40 50

600

620

640

660

08/11/2012 14

Control design. Bottom-up synthesis

DOC = DO Controller pHC = pH Controller LC = Level Controller RC = TNO2/TNH Controller NO2C = NO2 Controller

Supervisory layer

TNO2/TNH ratio

N2 conc. (Anammox reactor)

g/m

3 N2

Time (days)

Reponse to +5% step input in TNH input

The amount of N2 produced is maximized

More complex structure

The cascade takes advantage of the time-scale separation of the reactors

0 10 20 30 40 50550

600

650

700

CascadeNested cascade

0 10 20 30 40 501.1

1.12

1.14

1.16

Adaptation of microorganisms can be tackled

08/11/2012 15

Evaluation. Dynamic simulation

Benchmark simulation model 2 effluent

0 2 4 6 8 10 1250

100

150

200

250

300

350

Time (days)

Flow

rate

(m3 /d

)

0 2 4 6 8 10 121420

1430

1440

1450

1460

1470

1480

Amm

oniu

m co

ncen

trat

ion

(g/m

3 )

FlowrateTotal ammonium

0 2 4 6 8 10 12500

550

600

650

Time (days)

N2 c

once

ntra

tion

(g m

-3)

RegulatoryCascadeNested cascade)

0 2 4 6 8 10 127.5

8

8.5

9

Time (days)

pH

RegulatoryCascadeNested cascade)

Nitrogen production (Anammox) pH regulation (Anammox)

08/11/2012 16

Evaluation. Results

Structure Nitrogen removal DO SHARON pH SHARON pH Anammox

Regulatory 83.1% IAE 0.53 d 57.3 d 40.4 d

TV 908 d-1 5.40·10-4 m3d-1 1.60·10-5 m3d-1

Cascade 88.9% IAE 6.22 d 60.0 d 29.2 d

TV 9.51·103 d-1 2.90·10-4 m3 d-1 2.00·10-5 m3 d-1

Nested cascade 95.6%

IAE 7.23 d 60.0 d 29.3 d

TV 5.50·103 d-1 1.90·10-4 m3 d-1 1.90·10-5 m3 d-1

SP

0IAE= y -y dt

∞

∫n

i+1 ii=0

TV= u -u∑

08/11/2012 17

Conclusions

Due to the lack of actuators, selection of controlled variables is challenging in bioreactors

We used two methods to select the controlled variables:

- the Closed Loop Disturbance Gain, emphasizing the effect of disturbance rejection

- a trade-off between performance and robustness, through the synthesis of a H∞ controller

The regulatory layer was improved in order to address the process objectives, thereby producing designs of increasing complexity which can be adapted to the user’s needs

Incremental design of control system of SHARON-Anammox process for autotrophic

nitrogen removal

Funded by Danish Agency for Science, Technology and Innovation through the

Research Centre for Design of Microbial Communities in Membrane Bioreactors

(09-067230) for funding of the project

Miguel Mauricio-Iglesias Borja Valverde-Pérez Gürkan Sin