Industrial Experience at Lenzing AG and Petronor...1 Industrial Experience at Lenzing AG and...

1

Industrial Experienceat Lenzing AG and Petronor

Bernhard Voglauer

(Lenzing AG)

José Luis Pitarch

(UVA)

SPIRE Workshop

Jan. 27th 2016

2

Outline

SPIRE Workshop• Brussels27/01/2016

• Introduction Lenzing AG

• Overview Lenzing Case

• MORE Experiences at Lenzing

• Overview Petronor Case

• MORE Experiences at Petronor

• Summary

3

The Lenzing Group 2014

– Sales: EUR 1,864.2 mn (2013: 1,908.9 mn)

– Export share: 92.3% (2013: 90.8%)

– Fiber sales volumes: 960.000 tons (2013: 890,000 tons)

– Staff: 6.356 (2013: 6,675)

– Listed at Vienna Stock Exchange, Prime Market (ATX)

– Major shareholders:

• B&C Privatstiftung >50%

• Oberbank AG >5%

4

Our core market:Man-made cellulose fibers

– Produced from the raw material wood

• Halfway position between natural and chemical fibers

• Natural wearing properties of natural fibers combined with the advantagesof synthetical fibers such as purity and consistent quality

5

Lenzing/Austria2015

297.000 t Dissolving Pulp

337.000 t Cellulose Fibers

TENCEL®, Modal®, Lenzing Viscose®

6

Overview Lenzing Case

SPIRE Workshop• Brussels27/01/2016

Case 1.1.: Continuousoptimization of actualenergy consumption

• Optimize operation by the right settings of theavailable manipulated variables!

• Define REIs to be optimised

• DCS based solution preferred

• Fully automated

• Realized savings

300k€/a = 1.000.000 m3 nat gas/a = 2700 t CO2/a

7

Overview Lenzing Case

SPIRE Workshop• Brussels03/02/2016

• Definition of REIs

• Cleaning efficiency

• Modelling incl. fouling

• Optimizer Design

• Feasibility Study

• Savings: to be evaluated

Case 1.1.: Continuousoptimization of actualenergy consumption

Case 1.2.: Optimizationof cycle time

8

Overview Lenzing Case

SPIRE Workshop• Brussels03/02/2016

• Optimize specific steam consumption in totalevaporator network (30 Evaporators in 8 loops)

• Modelling, process control strategy, Optimizer,Visualisation, PIMS Implementation

• Savings: to be evaluated

Case 1.1.: Continuousoptimization of actualenergy consumption

Case 1.2.: Optimizationof cycle time

Case 2.1.: Continuousoptimization of actualenergy consumption(network)

9

Overview Lenzing Case

SPIRE Workshop• Brussels03/02/2016

Combination of all applied to the entire network:

• Which evaporator shall be operated at whichload in which loop best at optimized cycle time?

• Feasibility study

• Savings: to be evaluated

Case 1.1.: Continuousoptimization of actualenergy consumption

Case 1.2.: Optimizationof cycle time

Case 2.1.: Continuousoptimization of actualenergy consumption(network)

Case 2.2.: Optimizationof overall cycle cost ofevaporator network

10

What shall be optimized?• Think about influences and impacts, feasibility?

• use REI methodology MORE REI – Guideline!

Examples

MORE Experiences

03/02/2016 SPIRE Workshop• Brussels

NEW!

NEW!

11

Availability of information?• Suitable design of visualization MORE Visualization – Guideline!

• on suitable platform IntexcSuite, PIMS

• to the relevant people!

MORE Experiences

27/01/2016 SPIRE Workshop• Brussels

12

Make the right decision!• decision support system

• Implementation of model based optimization methods UVA, TUDO

Examples:

static grey box model Evaporator 38: UVA

fouling model: UVA

Optimization of one cycle: UVA

Feasibility study cycle optimization: UVA

static black box model Evaporator: TUDO

Optimization of actual steam consumption in evaporator network: TUDO

MORE Experiences

27/01/2016 SPIRE Workshop• Brussels

13

• First principles based stationary mathematical model

27/01/2016 SPIRE Workshop• Brussels

Grey Nonlinear Model (UVa)

Spinbath C-P-T equilibriums.

Heat transfer to ambient.

Psychrometric conditions. Energy and massbalances.

14

• Instantaneous RTO:

• Already implemented

Results for line EV.40

27/01/2016 SPIRE Workshop• Brussels

• Cleaning prediction:

SOLUTION: BIG CLEANING

Cleaning in 24 days, 16 h, Update control each 6hSolver time < 1 min (Intel® Core™ i7-4510 CPU)

0,2

0,22

0,24

0,26

0,28

0,3

0 100 200 300 400 500 600 700 800

Time (h)

Measured

Optimized

SOLUTION:Maximum spinbath temperatureMinimum circulating flow

Unitary cost (€/h)

Specific steam consumption

Specific steam consumption

15

• Largest site in Spain: production of 11 Mi T/yearBilbao

• 945 fixedemployees

• 6200 inducedjobs

Petronor: Oil refinery

27/01/2016 SPIRE Workshop• Brussels

16

Overview Petronor case

27/01/2016 SPIRE Workshop• Brussels

MOTIVATION

Hydrogen H2 is an expensive utility, very important in the refinery global economicbalance.

H2 requirements have experienced a steady increase in recent years:

o heavier fuels are processed;

o more strict environmental regulations;

H2 production capacity is a bottleneck for oil processing capacity.

H2 consumption can experience frequent significant changes (feedstock usuallychanges every 2-3 days)

GOAL: INCREASE EFFICIENCY IN H2

USE IN THE REFINERY

17

MORE Experiences

27/01/2016 SPIRE Workshop• Brussels

Definition and computation of REIs

REI 1. Intensity

REI 2. Efficiency

REI 3. Relative Efficiency (compared to best case)

o Reference is obtained by solving an optimization problem with a material balancemodel of the H2 network.

REI 4. Operational Efficiency (secondary REI with operational purposes)

REI 5. Controllability Efficiency (secondary REI with operational purposes)

o Use of REIs for Network on-line Decision Support purposes

ͳ�௦௨ܫܧ �௧�ூ௧௦௧௬

=ݎݑ ଶ�ܪ� ݑݏ � ݎ� �ሺݎݐ ଷ/ℎ)

ݕ ݎ ݎ ݎ� ݏݏ �ሺ ଷ/ℎ)

ͳ�ுଶ�ே௧ܫܧ�ூ௧௦௧௬

=∑ ݎݑ ݎ�ଶܪ� ݑ ௗ௨�௧௦ �ሺ ଷ/ℎ)

∑ ݕ ݎ ݎ ݎ� ݏݏ ௦௨ �௧௦ �ሺ ଷ/ℎ)

͵ܫܧ �ுଶ�ே௧ோா௬

=∑ ܯ�� ݑ � ݑݍ ݎ ݐܯ� ݎ � ݎݑ ݎ�ଶܪ� ݑ �ሺ Ǥሻௗ௨�௧௦

∑ ܣ�� ݑݐ ܯ� ݐ ݎ � ݎݑ ݎ�ଶܪ� ݑ ௗ௨�௧௦

1827/01/2016 SPIRE Workshop• Brussels

Optimal management of H2 network

Producer units: STEAM-REFORMING FURNACESH2 byproduct: CATALYTIC PLATFORMERSConsumer units: HDS, HDT

99.9%

96%

75%

Which plants must produce H2 and their Production rates?

19

MORE Experiences

27/01/2016 SPIRE Workshop• Brussels

CHALLENGES: an accurate plant state estimation is difficult.

Great variability for H2 consumption in HDT hydrotreating plants:

Lack of on-line measurements for H2 purity and gas molecular weight MW.

Error prone Orifice-Plate Differential Pressure flowmeter. Drift error due to:

Great influence of variability in light ends composition over total gas MW

o Uncertainty unless both H2 purity and gas MW are measured.

o PRECISE ONLINE COMPUTATION OF REI’s WAS NOT POSSIBLE

o EXPECTED BENEFITS around 2-4% in opportunity for profit inoperation due to frequent changes in scenarios

20

Data reconciliation is used to provide consistent estimates: Simplified solubility model to determine the item (MP/LP purges)

Disregard flowmeters with gross errors persistent in time.

Estimate H2 consumption in reactors according to FI/AI measurements.

Real-time optimisation tests to compute REI 3 (outline for the moment)

MORE Experiences

27/01/2016 SPIRE Workshop• Brussels

20,00

30,00

40,00

50,00

60,00

70,00

80,00

90,00

100,00

(%)

REI 1 - Intensity

HD3 - pure H2cons/Hydrocarbon

Min non-pure H2 / Actual non-pure H2

0,90

0,95

1,00

1,05

(%1

)

REI3 - Relative Efficiency

NET - Min non pure H2 prod/Actual non pure H2 prod

21

• A What-If analysis regarding REIs is available insimulation

MORE Experiences

27/01/2016 SPIRE Workshop• Brussels03/02/2016

decision variables u

Optimization(SNOPT/IPOPT)

SimulationEcosimPro

Process

J cost funtion

g constrains

Data Adquisition(SCADA + Excel )

Data Treatment(EcosimPro/

Excel)

SimulationEcosimPro

Sequential (EcosimPro) /Simultaneous (GAMS)

REIs for off-lineDecision Support

22

• A DMC controller has been commissioned and is in operation– Deals with the main source of inefficiency in the H2 network operation

– Complementary to the RTO approach

– Built upon other MPCs managing the operation in each consumer plant

MORE Experiences

27/01/2016 SPIRE Workshop• Brussels

23

Summary

27/01/2016 SPIRE Workshop• Brussels

• SUITABLE RESOURCE EFFICIENCY INDICATORS have been defined.

• SUITABLE CLOSED-LOOP SOLUTIONS have been developed and are alreadyonline and working!

• DECISION SUPPORT SYSTEMS have been designed and are in development.

• SAVINGS ARE ACHIEVED: LENZING: around 1.5 MioNm3/y of natural gas and300 k€/y, PETRONOR: around 1.5% high-purity H2 consumption)

… and more are expected!

• MORE IDEAS and EXCELLENT NETWORK OF PARTNERS responsible forachievements

24

End

27/01/2016 SPIRE Workshop• Brussels