CHEMICAL ENGINEERING TRANSACTIONS

VOL. 57, 2017

A publication of

The Italian Association

of Chemical Engineering
Online at www.aidic.it/cet

Guest Editors: Sauro Pierucci, Jiří Jaromír Klemeš, Laura Piazza, Serafim Bakalis
Copyright © 2017, AIDIC Servizi S.r.l.

ISBN 978-88-95608- 48-8; ISSN 2283-9216

Dynamic Simulation of Disturbances Triggering Loss of
Operability in a Biogas Production Plant

Abdulbari S. M. Salm, Valeria Casson Moreno, Giacomo Antonioni, Valerio
Cozzani
Alma Mater Studiorum - Università di Bologna, Dipartimento di Ingegneria Chimica, Mineraria e delle Tecnologie Ambientali,
via Terracini 28, 40131 Bologna, Italy
giacomo.antonioni3@unibo.it

In the recent years, the biogas industry is rapidly growing, but the sector has been affected by several major
accidents related to its supply chain, which pointed out the need to perform a detailed process safety analysis.
Hazards and operability analysis (HazOp) is one of the most used and efficient technique for the identification
of potential accidents, but its main limitation is the qualitative nature of the results. This study aims to override
this limitation by supporting HazOp study using a dynamic simulator, which is able to reproduce deviations
and operational failures in a typical biogas production plant, in order to help in the quantification of HazOp
results and to improve safety and reliability of the process.
The main result of the study is the maximum biogas production rate that can be safely handled by the system
under normal conditions and then the maximum rate that can be allowed if some safety barrier (though
correctly designed) should fail has been evaluated too. The most critical failure identified is a malfunctioning in
the pressure relief system of the digester, because under this condition, the maximum biogas production rate
that can be reached is about 2/3 of the corresponding value for a properly working system.

1. Introduction

Over the last 15 years, producing biogas from anaerobic digestion as a form of energy from renewable
sources has become more and more common in Europe and worldwide (EurObserv’ER, 2014; REN21, 2015).

For this reason, biogas sector experienced a fast growth, in terms of number of related facilities and installed
capacity: in 2014, in the sole Europe, nearly 8 GW of energy have been produced from biogas (REN21, 2015).
Unfortunately, this increasing trend has been accompanied by an even faster increase in the trend of major
accidents related to biogas supply chain (Casson Moreno et al., 2016; Casson Moreno and Cozzani, 2015).
The analysis of the accidents revealed several risk and operability issues with biogas production. From the
process safety stand point, the majority of biogas production plants are medium to small scale, therefore not
regulated by the Seveso III Directive (2012/18/EU); ATEX (2014/34/UE) and D.lgs 81/2008 are requirements
only. On the other side, anaerobic digestion for biogas production entails some complex biological reactions
(Findeisen, 2015), causing hazards and operability problems in facilities that usually are not managed by
personnel familiar with biotechnological processes and generally are poorly trained in process safety aspects
(Saracino et al., 2016). The overall outcome was the need for ad hoc hazard identification and operability
studies specific for biogas facilities.
There are several consolidated hazard identification techniques, such as checklists, Failure Mode and Effect
Analysis (FMEA), What-if analysis etc. Among them, Hazard and Operability study (HazOp) (Center for
Chemical Process Safety - CCPS, 2008). HazOp study is recognized as one of the most used and frequently
modified methodology for process safety analysis, but its main disadvantages are its time-consuming nature
(mainly due to different expertise required for its correct application) and the qualitative nature of the results
(Pasman and Rogers, 2016).
The progress in computer technology has created possibilities to eliminate or reduce the disadvantages of
conventional techniques for process hazard analysis (Dunjó et al. 2010, Janošovský et al. 2016). Process
simulation is hardly a new concept, but has been adopted in process engineering for decades. Traditionally,

DOI: 10.3303/CET1757100

Please cite this article as: Salm A.S.M., Casson Moreno V., Antonioni G., Cozzani V., 2017, Dynamic simulation of disturbances triggering loss
of operability in a biogas production plant, Chemical Engineering Transactions, 57, 595-600 DOI: 10.3303/CET1757100

595

simulation could be classified into two types, steady state simulation and dynamic simulation, with the latter
deemed as the more powerful and versatile engineering tool in process engineering and also for safety
examination, several examples for its use for study of operational failure of chemical process has been
documented (Pannocchia and Landucci, 2014, Ramzan and Witt, 2007).
The goal of the presented paper is to support HazOp analysis that has been performed on a typical biogas
production plant from anaerobic digestion, using dynamic simulation environment, including use of Aspen
HYSYS v8.6 process modelling to determine the effects of operational disturbances on the entire system and
its safe operating limits.

2. Methodology

Hazard and operability (HazOp) studies have become a significant part of the design of new process plants
and of the revision of existing plants in the process industry (Center for Chemical Process Safety - CCPS,
2008). The HazOp analysis systematically identifies all the possible causes and consequences within the
system for each hypothesized deviation of one of the variables of the process: the research is carried out
applying a set of qualitative guidewords (such as 'No', 'less', higher, 'instead') to the process variables of the
plant and determining all process variable deviations.
On the other hand, the design and optimization of a chemical process involves the study of both its steady
state and dynamic behaviour. While in steady state energy and material balances are solved to evaluate
different plant scenarios, with dynamic simulation, we can confirm that the plant can produce the desired
product in a manner that is safe and easy to operate. By defining detailed equipment specifications in the
dynamic simulation, we can verify that the equipment will function as expected in an actual plant situation.
With dynamic analysis, we can design and test a variety of control strategies before choosing the one that
may be suitable for implementation, and examine the dynamic response to system disturbances and optimize
the tuning of controllers.
The idea is to represent the plant and part of its control and safety system within a dynamic process simulator
(Aspen HYSYS), which should be able to simulate dynamically deviations previously identified in a HazOp
analysis, taking also into account the possibility of failure in single equipment or protection system. Thus
possible effects of these deviations can be rigorously quantified and operational safety limits identified.

3. Description of the reference scheme and simulation

The most typical configuration of a biogas production plant, which consists of a single digester with internal
desulfurization unit where the biogas is produced, has been selected as a reference case study (Figure 1).
The size of the plant and its operating conditions, reported in Table 1, are those typical of a 1 MW production
plant (Scarponi et al., 2016, 2015).

Figure 1. Block diagram of the biogas production plant used in the dynamic simulation.

HazOp analysis performed on this configuration, showed that the most important deviations in process
variables may lead, to an abnormal increase in the biogas production rate, which in turn causes an increase in
the pressure inside the digester (and in the downstream units), if equipment and/or protection systems should
fail, as schematically shown in Figure 2.

ANAEROBIC DIGESTER WITH
GASOMETER AND INTERNAL
DESULFURISATION SYSTEM

HYDRAULIC
SEAL

CHILLERCATCH
BASIN

BLOWER

FLARE

CHP

596

Table 1: Operating conditions for the biogas production plant used in the dynamic simulation.

Process variable Digester Blower Dryer
Flow rate [Nm3/h] 300 300 300
Biogas inventory [m3] 900 N.R. N.R.
Temperature [°C] 47 47 15
Pressure [kPag] 1 2 2
Composition [CH4/CO2] 60/40 60/40 60/40
H2S [ppm vol] 200 200 <100
Piping max. diameter [m] 0.150 0.150 0.150

Pressure increase should be limited through three different systems:
1. the basic pressure control acting on the valve before the Combined Heat and Power unit (CHP valve),
which should be automatically switched from 60 % to 100 % opening when the digester pressure is greater
than 1 kPag;
2. the flare, whose valve is completely opened when the pressure reaches 2 kPag;
3. the hydraulic seal directly installed on the vapour space of the digester, which starts relieving the gas at 4
kPag (maximum head of 6 kPag) in order to keep the pressure below the safety threshold of 8 kPag.

Figure 2: Schematic description of the analysis.

The plant and part of its control system has been reproduced within the simulator, by means of the scheme
shown in Figure 3. Since the biogas produced (B-G out) is saturated with water it passes through an
underground pipe, simulated as a cooler (E-102), to reduce the humidity. The condensate is separated in a
separator (Sep1) and dumped out using a pump (P-100) actuated by an on-off level controller, while the
biogas is further cooled in a chiller (E-100) with a temperature controller where the condensate is separated in
a second separator (Sep2) and dumped out through a valve (VLV-103) actuated by another on-off level
controller. The dry biogas is then sent to the CHP unit through a blower (K-100). The nominal operating
conditions for the production plant are presented in the Table 1.
The content of the digester is simulated using two streams (Gas_feed and water_in). Gas_feed stream is also
used to simulate the biogas production rate and can be freely manipulated thanks to the controller FIC-100.
Cooler E-101 is used to simulate the coil that recirculates water for heating purposes. The plant has three
level controllers to maintain the level inside the vessels (Digester, Sep1, and Sep2), and two temperature
controllers to maintain the temperature of the Digester and of the second separator (Sep2).
The first system to protect the digester against overpressure is simulated through PIC-100 controller, which
switches the CHP valve from 60 % to 100 %. The second measure is simulated by the on-off controller DIG-
100, which activates the flare valve (VLV-102) that sends the biogas to torch. The third measure is the
hydraulic seal, which is simulated as a relief valve (RV-100) with a set pressure of 4 kPag, a full open
pressure of 6 kPag, and an orifice area of 800 mm2, calculated in order to relief the nominal biogas production
rate when all the remaining systems should fail.
The plant will be analysed at two operational levels. At the first level (Case 1), it will be analysed without
failure in any equipment (CHP valve and protection systems), while at the second level (Case 2, Case 3, and
Case 4), failure of single equipment or protection system is considered.

597

In all the cases, the increase in biogas production is simulated through a ramp (after 20 minutes of normal
operation) in order to obtain more realistic results. The flow rate linearly increases from the nominal value to
the maximum allowable flow rate that can be handled by the system without explosion of the digester, which is
assumed to occur when the pressure inside the digester reaches the design value (8 kPag).

Figure 3: Process flow diagram of the plant representation within the simulator.

3.1 No failure in any equipment or protection system (Case 1)

The flow rate start increasing until it reaches 310 Nm3/h (after 5 minutes from the ramp start), when the CHP
is switched to 100 % opening. The flare (VLV-102) is activated after 11 minute, when the flow rate reaches
353 Nm3/h and the relief valve (RV-100) starts opening after 19 minute, at flow rate of 437 Nm3/h, until it
becomes fully open at a flow rate of 779 Nm3/h, after 51 minute. The maximum allowable value for the biogas
flow rate is found to be 886 Nm3/h (see Figure 4a) because the digester pressure floats around 8 kPag but
does not exceed this limit value (Figure 4b).

Figure 4: Results of Case 1: a) biogas production rate, b) digester pressure, c) valves opening, d) liquid level

in Sep1.

200

400

600

800

1000

0 50 100 150 200

S
td

.
G

a
s
F

lo
w

(
N

m
3
/h

)

Time (min)

(a)

0 50 100 150 200
Time (min)

P
re

s
s

u
re

(
K

P
a

g
)

(b)

0
20
40
60
80

100
120

0 50 100 150 200

V
a

lv
e

o
p

e
n

in
g

Time (min)

(c)

CHP
VLV-102
RV-100

0 50 100 150 200

L
iq

u
id

l
e

v
e

l
%

Time (min)

(d)

598

3.2 Failure in one equipment or protection system (Cases 2 to 4)

The same disturbance (ramp increase in the biogas production rate) has been applied to other three situations
(summarized in Table 2), in which one item has been considered has failed during the deviation.

Table 2: Summary of the main result of all cases studied.

Item
Case 1 Case 2 Case 3 Case 4

Flow rate
(Nm3/h)

Time
(min)

Flow rate
(Nm3/h)

Time
(min)

Flow rate
(Nm3/h)

Time
(min)

Flow rate
(Nm3/h)

Time
(min)

CHP 310 5 Failure 310 5 310 5
Flare 353 11 activated manually Failure 353 11

RV-100 (start open) 437 39 364 14 418 18 Failure
max. allowable flow rate 886 n.r. 660 45 850 60 602 35

From the analysis of the time at which 8 kPag is reached (Time To Failure of the digester) reported in the last
row of Table 2, it is clear that the most critical situation is when the relieving system (hydraulic seal) is not
available (Case 4), because the time required before digester failure is only 35 minutes.
Under these conditions, the CHP is switched to 100 % opening after 5 min, as for Case 1. The flare (VLV-102)
is activated after 11 minute but, since the hydraulic seal fails to let the gas flow out, when the flow rate
reaches the value of 602 Nm3/h (after 35 minute), the pressure increases up to 8 kPag and over causing the
failure of the digester (see Figure 5).
Similar conclusion with different Time to Failure can be obtained for Cases 2 and 3, being the latter,
reproducing a failure of the flare, the less critical because the system can withstand up to 1 hour of biogas flow
rate increase.

Figure 5: Results of the Case 4 "failure in relief system" (a-biogas production rate, b-digester pressure, c-

valve opening).

4. Conclusions

The results of the plant analysis show that when all equipment and safety barriers are properly working, the
system could handle about three times the nominal biogas production rate and the system reaches a sort of
steady state within the safe operating limits, i.e. below the design pressure of the digester (explosion
pressure). When a failure in a single equipment or safety barrier is considered, the most critical situation is
when the hydraulic seal is out of service, because the system could handle only about two times the nominal
biogas production rate before reaching the design pressure of the digester after approximately 35 minute.

200

400

600

800

1000

0 20 40 60 80 100

S
td

.
G

a
s

F
lo

w
(

N
m

3
/h

)

Time (min)

(a) biogas production Explosion limit

0 20 40 60 80 100

P
re

s
s

u
re

(
K

P
a

g
)

Time (min)

(b) Digester pressure Explosion limit

100

150

0 20 40 60 80 100

V
a

lv
e

o
p

e
n

in
g

Time (min)

599

When the CHP is stopped (or its valve failed closed) the flare should be manually opened (e.g. during
maintenance) and the system can handle about 2.3 times the nominal biogas production rate before reaching
the design pressure of the digester after approximately 45 minutes. A failure in the flare system seems to be
safer, since the system is still able to handle about three times the nominal production rate, even if after 1 hour
the digester failure pressure will be anyway reached.
As a general result, this study showed that dynamic simulation is a powerful tool to analyse disturbances and
operational failures in a process plant, being in the quantification of deviations only qualitatively identified in a
typical HazOp study.

References

Casson Moreno V., Cozzani V., 2015. Major accident hazard in bioenergy production. J. Loss Prev. Process
Ind. 35, 135–144. doi:10.1016/j.jlp.2015.04.004

Casson Moreno V., Papasidero S., Scarponi G.E., Guglielmi D., Cozzani V., 2016. Analysis of accidents in
biogas production and upgrading. Renew. Energy 96, 1127–1134. doi:10.1016/j.renene.2015.10.017

Center for Chemical Process Safety - CCPS, 2008. Guidelines for Hazard Evaluation Procedures, 3rd Edition.
ed. AIChE, New Jersey.

Dunjó J., Fthenakis V., Vílchez J. A., Arnaldos J., 2010, Hazard and operability (HAZOP) analysis. A literature
review, J. Hazard. Mater. 173, 19-32.

Janošovský J., Labovský J., Jelemenský L., 2016, Automated Model-based HAZOP Study in Process Hazard
Analysis, Chemical Engineering Transactions, 48, DOI: 10.3303/CET1648085.

Pannocchia G., Landucci G., 2014, On the use of dynamic process simulators for the quantitative assessment
of industrial accidents. Chemical Engineering Transactions, 36, 505-510. DOI: 10.3303/CET1436085

Pasman H.J., Rogers W.J., 2016. How can we improve HAZOP, our old work horse, and do more with its
results? An overview of recent developments. Chem. Eng. Trans. 48, 829–834. doi:10.3303/CET1648139

Ramzan N., Witt W., 2007, Combining disturbance simulation and safety analysis techniques for improvement
of process safety and reliability, Elsevier B.V.

REN21, 2015. Renewables 2015 Global Status Report.
Saracino A., Casson Moreno V., Antonioni G., Spadoni G., Cozzani V., 2016. Application of a self-assessment

methodology for occupational safety to biogas industry. Chem. Eng. Trans. 53, 247–252.
doi:10.3303/CET1653042

Scarponi, G.E., Guglielmi, D., Casson Moreno, V., Cozzani, V., 2016. Assessment of inherently safer
alternatives in biogas production and upgrading. AIChE J. 62. doi:10.1002/aic.15224

Scarponi, G.E., Guglielmi, D., Casson Moreno, V., Cozzani, V., 2015. Risk Assessment of a Biogas
Production and Upgrading Plant. Chem. Eng. Trans. 43, 1921–1926. doi:10.3303/CET1543321

600