Accepted Manuscript

Thermodynamic analysis of a CNG refueling station considering the reciprocating

compressor

Morteza Saadat-Targhi, Javad Khadem, Mahmood Farzaneh-Gord

PII: S1875-5100(16)30027-0
DOI: 10.1016/j.jngse.2016.01.027
Reference: JNGSE 1231

To appear in: Journal of Natural Gas Science and Engineering

Received Date: 31 October 2015

Revised Date: 16 January 2016
Accepted Date: 21 January 2016

Please cite this article as: Saadat-Targhi, M., Khadem, J., Farzaneh-Gord, M., Thermodynamic analysis
of a CNG refueling station considering the reciprocating compressor, Journal of Natural Gas Science &
Engineering (2016), doi: 10.1016/j.jngse.2016.01.027.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Thermodynamic analysis of a CNG refueling station considering the

reciprocating compressor

Morteza Saadat-Targhia*, Javad Khadema , Mahmood Farzaneh-Gordb

a
The Faculty of Mechanical Engineering, University of Birjand, Birjand, Iran

PT
b
The Faculty of Mechanical Engineering, Shahrood University of Technology, Shahrood, Iran

* [email protected], Tel-Fax:+98(561)2502133

RI
Abstract

SC
The reciprocating compressor is the heart of a CNG refueling station with dominating influence on refueling
process. The main operating cost of a CNG refueling station arises from the compressor input work. The
reciprocating compressor have not been subjected to previous studies which investigated the CNG refuelling
stations. The main objective of the present work is to apply a thermodynamic analysis for modeling a CNG refueling

U
station, including on-board NGV cylinder, reservoir tanks, connecting pipes and espesially reciprocating
compressor. Unlike most previous researches, pressure variation of the reservoir tanks during the filling process
AN
have been considered which necessary for compressor modeling. The results of the present work have been
validated against the previous experimental data and theoretical studies. The additional experimental data, measured
by the present authors, is also used to analyze and identify the compressor performance in unsteady conditions. The
simulation result of mass flow rate is a good agreement with the experimental values. The results show that
M

variations in the compressor discharge pressure have no effect on the first and intermediate compressor pressure. It
has also no effect on the compressor delivery mass flow rate. The energy consumption by the compressor for a cycle
(including discharge and charge zones) is found to be 61.86 kWh. The average specific energy consumption to fill a
D

vehicle is found to be 0.250 kWh/kg.

TE

Key words: Thermodynamic modeling - CNG refueling station - Fast filling process - Reciprocating compressor

1. Introduction
EP

Compressed natural gas (CNG) and compressed hydrogen gas (CHG) could be considered clean compared to
petrol or diesel. Natural gas primarily consists of methane, but also contains other hydrocarbons such as butane,
propane and ethane. CNG is stored in high-pressure cylinder on the vehicle (190 to 250 bar). Natural gas vehicles
C

(NGVs) connect to the high-pressure reservoir tanks (200 to 235 bar) at refueling stations during filling process.
Two main goals could be seen for refueling stations, operating costs at the refueling stations and the time to refuel a
AC

vehicle (filling time). The main operating cost of any CNG refueling station belongs to the compressor. So, a
simulation to optimize the refueling process is very important. Filling time less than 5 minutes is referred as a fast
fill. In these refueling stations, CNG is stored in reservoir tanks. There are two systems for storing CNG in reservoir
tanks, cascade (multi-line) and buffer (mono-line) storage systems. Reservoir tanks in a cascade storage system are
divided into three pressure levels, low, medium and high pressures. In a buffer system, reservoir tanks are only in
one pressure level. A refueling station with a cascade storage system uses the priority panel to control the outlet flow
of compressor to a series of cascade storage tanks (see Fig. 1). Based on the algorithm in the dispenser the on-board
NGV cylinder is switched from lower pressure reservoir to a higher one, when the mass flow rate drops to a certain
limit. There are many similarities between the CNG and CHG refueling process which makes the present work as a
useful research for the CHG refueling stations.

1
 ACCEPTED MANUSCRIPT

There are various researches about CNG refueling station. These researches could be classified into 4 main
groups as: a) the researches which only consider on-board NGV’s cylinder, b) the researches which consider the
reservoir tanks and on-board NGV’s cylinder, c) the researches which consider the reservoir tanks, on-board NGV’s
cylinder and connection pipes and finally d) the researches which consider the reservoir tanks, on-board NGV’s
cylinder, connection pipes and reciprocating compressor.

Three have been many researches which only considered on-baord NGV’s cylinder. In these researches, the

PT
reservoir tanks modeled as a constant properties reservoir. Firstly, Kountz [1], presented a simple thermodynamic
model to modeling the fast filling process. The model was simple with only considering the on-board NGV cylinder.
The heat transfer from the on-board NGV cylinder to the environment is investigated in his work. Kountz et al. [2-
5], have also developed a dispenser control algorithm. Farzaneh-Gord et al. [6] modeled the process of fast filling

RI
for one storage system (buffer) in a CNG refueling station. The effects of ambient condition and initial reservoir
tanks pressure on final on-board NGV cylinder pressure, temperature and filling time are considered in their works.
Their model based on the on-board NGV cylinder, without considering the connection pipes. They modeled fast

SC
filling process while the pressure of reservoir tanks, remain constant during the process. Farzaneh-Gord et al. [7]
presented a thermodynamic analysis of cascade storage bank. Their model was similar to previous models. The
results of this study indicated that ambient temperature and initial conditions have an enormous effect on final on-
board NGV cylinder conditions and filling time and mass. Their results indicated that there was a temperature rise of

U
at least 40 K for real gas through the fast filling process.They have also developed a theoretical analysis based on
real gas assumptions to compare two storage systems. The results show both of storage systems have advantages
AN
over the other. It has been found that the entropy generation in the cascade storage system is 50% less than the
buffer storage system. They presented, the less entropy generation which probably cases lower compressor input
work, but energy consumption by the compressor has not been calculated, directly. To calculate the energy
M

consumption by the compressor, it should be modeled. They showed that filling time in the buffer storage system is
about 66% less than the other. Their model based on two cylinders without considering pipe, similar to the former
ones. In a study, Bang et al, [8] has modeled a CNG residential refueling system. Their model was able to generate
electric power consumption profiles versus refueling time. The model is appropriate for simulating residential
D

refueling systems (filling time is about 4 hours) rather than public systems (fast fill). A theoretical analysis of an
adiabatic system, has been developed to study the effects of the natural gas compositions on the performance of
TE

refueling stations by Farzaneh-Gord et al. [9]. The results showed that the compositions of natural gas have great
effects on the final in-cylinder conditions at the end of the process. Furthermore, the gas with less methane
percentage in its composition is more suitable for faster filling process. For example, the filling time for Pars natural
gas type (with about 87% methane) is about 15 seconds less than Khangiran natural gas (with about 98% methane).
EP

Deymi-Dashtebayaz et al. [10] presented an equation for foretelling heat transfer rate in on-board NGV cylinder
based on a numerical simulation using the CFD code fluent. By using this equation and a thermodynamic model, the
final conditions of the on-board NGV cylinder at the end of the filling process, were determined. Their results had
C

been compared with an experimental research and showed a good agreement. There are similar studies which only
consider the on-board vehicle cylinder for CHG [11-20].
AC

To the author knowledge, the limited researches have been carried out which considered the reservoir tanks and
on-board NGV’s cylinder. Deymi-Dashtebayaz et al. [21] studied the effects of charging dynamic reservoir tanks
conditions on NGV cylinder filling time. Their results show that the initial pressure of reservoir tanks has a big
effect on the reservoir tanks volume for bringing up the NGV cylinder to 200 bar. Failure to correctly predict the
mass flow rate is evident in their results. The reason is the lack of connection pipes modeling. A thermodynamic
model for a tank filling process is presented for CHG by Striednig et al. [22]. Their simulation is used to predict the
hydrogen properties during the fast filling process. They considered the reservoir tanks and on-board NGV’s
cylinder in their model but the mass flow rate hasn’t been calculated and it has been assumed constant during filling.

2
 ACCEPTED MANUSCRIPT

There is also only one research which considered the reservoir tanks, on-board NGV’s cylinder and connection
pipes. A mathematical model has been developed to calculate mass flow rate of CNG considering connecting pipe
by Khadem et al. [23]. The natural gas flow in connecting pipe was modeled as a Fanno flow (ideal and adiabatic
flow where the effect of friction is considered) by them. In their work, the reservoir tanks modeled as a constant
properties reservoir, similar to many of the previous research. In their work, the mass flow rate of gas was calculated
by a trial and error procedure.

PT
Having a complete model for a CNG refueling station is of paramount importance. It should be noted that in all
previous work [1, 6-7,9-23], only a few components of the refueling station is modeled. The above literature review
shows that, there have not been any studies which modeled the reservoir tanks, on-board NGV’s cylinder,
connection pipes and reciprocating compressor simultaneously. The reciprocating compressor is the heart of any

RI
CNG refueling station, which has dominated influence on refueling process. Reciprocating compressors are used in
a lot of industrial applications. They can provide high pressure ratios [24]. There are a number of methods have been
developed for these tapes of compressors thermodynamic analysis [25-28]. There isn't any report about reciprocating

SC
compressor performance in unsteady conditions, too.

The main objective of the present study is to apply a thermodynamic analysis for a complete modeling of a CNG
refueling station. The natural gas flow between reservoir tanks and on-board NGV cylinder is modeled as a one-

U
dimensional and isothermal flow in the present work. the results of the present work have been validated against the
experimental data and compared with the results of a previous work [7], to investigate the effect of connecting pipes.
AN
On the other hand, unlike many of the previous researches [1, 6-7,9-23], pressure variations of the reservoir tanks
during the filling process have been considered in the present research. It is necessary for compressor modeling. The
reciprocating compressor of CNG refueling station has been modeled in the present work. The experimental data,
M

measured by the present authors, is used to identify the compressor performance in unsteady conditions.
D

2. Theory and modeling

TE

A CNG refueling station could be divided into six components (Fig. 1); compressor, priority panel, reservoir
tanks, connecting pipe, dispenser and on- board NGV cylinder. A simplified analysis of a CNG refueling station
could be performed by modeling these components. The compressor is used to fill the reservoir tanks, when
reservoir tank pressure is decreased to a certain limit. In a cascade storage system, reservoir tanks are filled on a
EP

priority identified by an algorithm. Based on another algorithm, the on-board NGV cylinder is switched from the
low pressure reservoir to the higher one, when the mass flow rate drops to a certain limit (in dispenser). Reservoir
tanks in a cascade storage system are divided into high, medium and low pressures (HP, MP and LP).
C

The theory used to model a CNG refueling station has been described in this section. The section involves four
AC

parts; on-board NGV cylinder, reservoir tanks, connecting pipe and compressor.

3
 ACCEPTED MANUSCRIPT

PT
RI
Fig. 1 A Schematic diagram of a CNG refueling station with cascade storage
system

U SC
2.1. On-board NGV cylinder AN
The on-board NGV cylinder is assumed as a simple thermodynamic open system. The system has a uniform
pressure and temperature. Applying the mass conservation principle to the control volume:

 




(1)
M

where  is inlet mass flow rate, described in the following section. The general form of the first law of
thermodynamics yields:
D


  ∑    ⁄2    
∑     ⁄2      ⁄    ⁄2  
  

TE

(2)

The change in kinetic and potential energy and the work term could be neglected as there is only one inlet,
yields:
EP

 
    
"


 !

 
(3)
C

The above equation could be more simplified by replacing enthalpy of the reservoir tanks:

 
   # 
AC




(4)

To simplify, the heat transfer of the on-board NGV cylinder to the environment is negligible. For an ideal gas
behavior:

  $% & ,   $ & , (  )& (5)

4
 ACCEPTED MANUSCRIPT

where $% and $ are constant pressure specific heats and constant volume specific heats, respectively. Knowing
that the reservoir tank temperature, the on-board NGV cylinder volume and specific heats are constant, equation (4)
could be simplified as follows:

* . $ &
/   $% &#

+ 

 ,-
(6)

PT
Finally, the first law of thermodynamics could be written as for the on-board NGV cylinder:

(
   &# *

0 ,
/




1 
(7)

RI
Using equation (7), the on-board NGV cylinder pressure in the new time step could be calculated. Using
equation (1) and the on-board NGV cylinder pressure in the new time step and ideal gas equation of state, the on-

SC
board NGV cylinder temperature in the new time step could be detected.

2.2. Reservoir tanks

U
Similar to the previous section, applying the mass conservation equation to the new control volume (reservoir
tanks) yields:
AN
 −

2


(8)

That  is outlet mass flow rate, described in the following section. General form of the first law of
M

thermodynamics for the new control volume (equation 2) yields:

 
 −  # 

2
D


(9)
TE

To simplify, the heat transfer of reservoir tanks to the environment is negligible. Similar to the on-board NGV
cylinder, the first law of thermodynamics could be written as for the reservoir tanks:

(#    &# *

0 ,
/


EP


1 2
(10)

Using equation (10), the reservoir tank pressure in the new time step could be calculated.
C

2.3. Connecting pipe

AC

The natural gas flow between reservoir tanks and on-board NGV cylinder is modeled as a one-dimensional and
isothermal flow. Transient process is a quasi-steady process. To calculate the mass flow rate of gas through the
connecting pipe, continuity, momentum and equation of state must be solved for a pipe.

The governing equations (Continuity, momentum and state equation) for pipe flow are as follows, [29]:

4
0
5
(11)

5
 ACCEPTED MANUSCRIPT

4 ( 7
  4   0
5 5 28
(12)

(  4 &
,
9 (13)

where, , ρ, V, P, f, D, R, M and T are compressibility factor, density, velocity, pressure, Fanning friction factor,

PT
diameter of pipe, gas constant, molecular weight and temperature, respectively.

Applying the mass Continuity equation:



RI
4 
:
(14)

SC
The combination of equations (12) and (14) gives:

75
4
;<
=
<>  4(
@  ;<=<>  0
2: 8
? ??
(15)

U
???
AN
The kinetic energy term (I) is negligible when compared with the other terms (II and III).

To integrate term (II) of the previous equation and by using equation (11):
M

+" +"
C( C +"
C( − (G 
A 4(  A (  A (( 
D)& DEF )&EF +B 2DEF )&EF
(16)
+B +B
D

Which [30]:
TE

&G  & 2 (G (
&EF  HI (EF  K(G  ( − L
2 3 (G  (
(17)

DEF , could be calculated based on the &EF and (EF for pure methane by an equation of state (Soave-Redlich-
EP

Kwong equation of state). This equation is convenient for thermodynamic modeling. So DEF is given by the Soave-
Redlich-Kwong equation of state [31]:
C

D M − D   :∗ − O ∗ 1  O ∗ D − :∗ O∗  0 (18)

AC

With, :∗ and O∗ are obtained by pressure and temperature, [31].

To integrate term (III) of the equation (15):

Q"
7 7 7
A 5   5 − 5G    R
2: 8 2: 8 2: 8
(19)

QB

6
 ACCEPTED MANUSCRIPT

where L is the length of pipe between reservoir tanks and on-board NGV cylinder. Substituting equations (16)
and (19) in equation (15), mass flow rate is given by:

1 C8(G − ( 
V.W
  : T U
S7 R)&EF DEF
(20)

By Spitzglass (medium pressure) equation [32]:

PT
1 88.5 V.W
X ]
1  0.09144⁄8  1.18118
(21)
S7

RI
So, mass flow rate could be obtained by equations (17), (18), (20) and (21).

SC
2.4. Compressor

The compressor type used in a CNG refueling station is reciprocating compressor (usually multi-stage). This

U
type of compressor use piston-cylinder combinations to compress gas. Fig. 2 shows the schematic diagram of three-
stage compressor with inter cooler between stages and after cooler at the outlet of the compressor. The most
AN
important parameters in compressor modeling are outlet mass flow rate and actual work of the compressor.
M
D
TE
EP

Fig. 2 Schematic diagram of three-stage reciprocating compressor

C
AC

In the next section (Analysis and discussion) it will be shown that the outlet mass flow rate of compressor
remains constant during the operation of the compressor based on experimental measurements. The outlet mass flow
rate of the compressor is calculated by the following equation:

CaFE_

^  4_
` _
 X ] ` 4E#,_
` _

CaE#
(22)

where Mwdef , Mwegh , ρegh,fjk and Q fjk are molecular weight of gas (Methane in the present work), molecular
weight of air, air density at standard conditions and the capacity of the compressor at standard conditions,
respectively.

7
 ACCEPTED MANUSCRIPT

The adiabatic compressor work is calculated by the following equation [33]:

m (
_ opG
E
  D_n &_n ) K  o − 1L
m−1 (_n
(23)

Total adiabatic compressor work for three stages is calculated by:

oB pG o" pG
mG (
_G m (
_

PT
oB o"
E
  D & ) qX ] − 1r  D & ) qX ] − 1r
(24)
mG − 1 _nG _nG (_nG m − 1 _n _n (_n
os pG
mM (
_M os
D_nM &_nM ) qX

RI
 ] − 1r
mM − 1 (_nM

SC
The actual compressor work is calculated by the following equation:

E

E 
t
` t
(25)

U
where ηv and ηw are compression efficiency and mechanical efficiency, respectively. These parameters are
AN
commonly reported by compressors companies.
M

3. Results and discussion

D

3.1. Validation
TE

The comparison between the results of present work, a previous work [7] and experimental values of George
[34] are presented in this section. The initial pressure of the on-board NGV cylinder is 72 bar and the cylinder
volume is 52 litters. There is a cascade storage system in experimental procedure. The initial pressure of the
reservoir tanks is 251 bar. The mass flow rate varies between 0.29 and 0.0 kg/s and when it drops to 0.04 kg/s, the
EP

on-board NGV cylinder is switched to the other reservoir tanks by dispenser algorithm. Fig. 3 shows the mass flow
rate vs. time. Fast filling process has started after 24 seconds and has ended after 36 seconds from the start time. It

reason of peak at time~55 is increased upstream pressure by switching to another reservoir tank. A good agreement
has started with exhausting the reservoir tank 1 and after 26 seconds, has switched to another reservoir tank. The
C

could be seen between experimental and the simulation values of the present work. The deviation may be due to the
considered assumptions. The simulation values show the filling time have only 2 second differences from
AC

experimental values. The difference is about 10 seconds for the previous work [7]. It should be noted that, in the
previous work [7], to determine the mass flow rate between the reservoir tanks and on-board NGV cylinder, the
ideal mass flow rate is multiplied by a discharge coefficient. The discharge coefficient is usually selected using
matching experimental and numerical values.

8
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
Fig. 3 Comparison between the simulations and experimental results for mass flow rate

In Fig. 4, the on-board NGV cylinder pressure variations pending fast filling process are shown for the
experiment, present and the previous work [7]. Note from the figure, the on-board NGV cylinder pressure for the
M

experimental work is less than the present work and previous works. It may be caused by the assumption of ideal gas
and adiabatic model in the on-board NGV cylinder. So, the results of the present work are more accurate than the
previous works.
D
TE
C EP
AC

Fig. 4 Comparison between the simulations and experimental values for pressure variations of NGV cylinder

9
 ACCEPTED MANUSCRIPT

3.2. Experimental analysis of compressor

As mentioned, the compressor type used in CNG refueling stations is reciprocating compressor. Such as NGV
cylinder and reservoir tanks, compressor performance is in unsteady conditions. The third stage output pressure
changes with time caused unstable conditions for compressor. Two challenges appeared when reciprocating

PT
compressor performance in unsteady conditions was considered. First: how does pressure distribution change with
changes in the output pressure of the third stage? Pressure distributions include inlet pressure compressor, the first
stage output pressure and second stage output pressure (Fig. 2). Second: how does mass flow rate change with
changes in the output pressure of the third stage?

RI
The experiments have been carried out on a typical CNG refueling station (Baghdar CNG station No.1043,
Mashhad, Iran). The CNG refueling station has 4 dispensers, 8 refueling nozzles and one compressor. The main

SC
characteristics data for the compressor are given in table 1.

Table 1. Characteristics of compressor used for experiments

Number of Stages

U 3
AN
Cylinder Double Acting / Single Acting Single Acting
Cylinder Bore 1st Stage 110 mm
Cylinder Bore 2st Stage 66 mm
M

Cylinder Bore 3st Stage 42 mm

Stroke 1st Stage 127 mm
Stroke 2st Stage 127 mm
D

Stroke 3st Stage 127 mm

Capacity 1500 Nm3/hr
TE

Compressor Shaft Speed 800 rpm

Gas Methane
Compression efficiency 0.85
Mechanical efficiency 0.95
C EP

The schematic of the experimental setup is shown in Fig. 5. The pressure sensors mounted externally (PT) shown
in the figure. Some locations of compressor (four points) and reservoir tanks (three points) are instrumented with
AC

pressure sensors to monitor the pressure during the variation. Inlet pressure changes during the time could be
detected by a pressure sensor located at compressor inlet. The first, second and third stage output pressure changes
could be detected by using the other pressure sensors located on the compressor. Reservoir tanks pressure changes
could be observed by using the sensors installed on them. Experimental procedure, measured by present authors has
been carried out for a period of time to determine the compressor performance in unsteady conditions. Compressor
performance is unsteady because the compressor outlet pressure changes during the time.

10
 ACCEPTED MANUSCRIPT

PT
RI
SC
Fig. 5 Schematic of the experimental procedure

The pressure distributions of the gas vs. time curve for the compressor are shown in Fig. 6. It could be realized

U
that the compressor has started after 6 minutes and has stopped after 35 minutes from start time. It is evident that the
compressor inlet pressure during the operation is always constant, about 15 bar. This value is the inlet gas pipeline
AN
pressure. The output pressure of the first stage and second stage, during the compressor operation are also almost
constant, about 42 and 88 bar respectively. The output flow of the third stage, after passing into the priority panel,
enters to reservoir tanks. According to the increase in pressure vs. time in reservoir tanks, the output pressure of the
third stage increases, too. Also, the output pressure of the third stage decrease because of the empty tanks. So, unlike
M

in previous cases, the output pressure of the third stage varies with time. This is evident in the figure. Therefore, it is
concluded that, pressure variations of the third stage, have no effect on the inlet and intermediate stage pressure. As
is shown in this figure, the natural gas is compressed in the first stage to 42 bar, then transferred to the second stage
D

and compressed to 88 bar, then transferred to the third stage of compression where the natural gas is finally
compressed to a pressure of reservoir tanks. This important consequence is the answer for the first challenge
TE

expressed previously.
C EP
AC

11
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
Fig. 6 Pressure measurement and monitoring of the compressor

Pressure versus time curve is shown in Fig. 7 for HP reservoir tank. As is shown in this figure, since the
beginning until 6 minutes, the pressure is reduced. The reason for this decrease is related to the increase in NGV
M

cylinder pressure. The compressor has started after 6 minutes and the high pressure reservoir tank is selected by
priority panel, immediately. The pressure is increased from 206 bar to 230 bar by compressor during the 30 seconds.
Pressure is continuously increased to 30 seconds. Therefore, Output pressure of the third stage also increases
D

continuously in 30 seconds. After the pressure was increased to 230 bar, the priority panel selects another tank to
fill. So, the pressure of the HP reservoir tank is reduced, because the gas comes out to the dispenser. The process of
TE

filling and empty the reservoirs continue until the compressor is turned off again. As is shown in this figure growth
rate for pressure is nearly constant during the 30 seconds and this growth rate is nearly constant at other times, too.
Pressure growth rate could be considered equivalent to the mass flow rate changes. Therefore, it is concluded that,
pressure variations of the third stage, have no effect on the mass flow rate of the compressor. This important
EP

consequence is the answer for the second challenge expressed previously.

C
AC

12
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M

Fig. 7 Pressure variations of HP reservoir tank during the time

D

3.3. Simulation results

TE

Natural gas primarily consists of methane, so in this work CNG was assumed pure methane. The thermodynamic
model is used to predict the dynamic pressure variation during fast filling process for a cascade storage system.
Information about CNG refueling station and the NGV is presented in table 2. It is obtained from a typical CNG
EP

refueling station (Baghdar CNG station No.1043, Mashhad, Iran). The following assumptions have been assumed
for extracting the results:
C

• The initial pressure of each NGV cylinder is 1 bar.

• Environment Temperature remains constant (293 K).
AC

• The final pressure of NGV cylinder is 200 bar, at least.

• Time for switching between vehicles is one minute.
• There is only one dispenser in the CNG refueling station.
• During the fast filling process, the compressor is off.
• Initial pressure and temperature for cascade storage system are 235 bar and 293 K, respectively.

Table 2. Some information about CNG refueling station and NGV on-board cylinder

Storage capacity of on-board NGV cylinder 101 lit

13
 ACCEPTED MANUSCRIPT

HP reservoir tank capacity 1920 lit

MP reservoir tank capacity 1920 lit
LP reservoir tank capacity 1920 lit
Priority panel 3 lines
Length of pipe 50 m

PT
Inside diameter of pipe 0.009 m
Initial pressure of on-board NGV cylinder 1 bar
Initial pressure of HP, MP and LP reservoir tanks 235 bar

RI
SC
Variations of pressure during the time for the cascade storage system are shown in Fig. 8. It should be noted that
a total of 26 vehicles could be filled without the compressor having to recharge the reservoir tank. After 105 minutes
(discharge time) the HP reservoir tank pressure has dropped down to nearly 200 bar. Cascade storage reservoir
recovery time is about 17 minutes (charge time). A complete cycle (discharge and charge) for a cascade storage

U
system takes about 122 minutes. For the first vehicle the reservoir tank pressure is 235 bar, after the first vehicle is
filled, the LP reservoir tank pressure will be 224 bar. The first three vehicles could be filled from the LP reservoir
AN
tank without having to switch to the next reservoir tank. The fourth vehicle could not be filled to the 200 bar level
from the LP reservoir tank. The fourth vehicle will initially be filled from the LP reservoir tank until the mass flow
rate drops to a certain limit. The MP reservoir tank will now have to be used to fill the vehicle. The twelfth vehicle
M

could not be filled to the 200 bar level from the LP and MP reservoir tanks. The HP reservoir tank will now have to
be used to fill the vehicle. There is an algorithm in the dispenser to manage the filling of NGV cylinders (discharge
zone). This algorithm applied to fill process from a cascade storage system. According to the algorithm, a NGV will
first be filled from the low pressure reservoir tanks, which will result an incomplete fill, perhaps to 220 bar. The
D

medium and high pressures reservoir tanks will bring the vehicle's tank closer to 220 bar. If this algorithm is
continued, 26 vehicles could be filled before cascade system will need to be charged. The average filling time for a
TE

cascade storage system, is about 242 seconds.

There is another algorithm in the priority panel to manage the filling of reservoir tanks which means that one
reservoir tank is filled after another (charge zone). Firstly, HP reservoir tank, then MP reservoir tank and finally, LP
EP

reservoir tank are filled by the compressor.

C
AC

14
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
Fig. 8 Pressure profile of reservoir tanks for the cascade storage system
M

To see more details, the mass flow rate of the gas for cascade storage system is shown in Fig. 9. The results are
presented for the first, tenth and twenty-fifth vehicles. Based on dispenser algorithm, the on-board NGV cylinder is
switched from lower pressure reservoir to a higher one, when the mass flow rate drops to certain limit (0.02 kg/s for
present work). According to the figure, the mass flow rate of gas profile for the tenth and twenty-fifth vehicles are
D

divided into two and three separate parts, respectively. As shown in the figure, filling time for the first, tenth and
twenty-fifth vehicle are 110, 155 and 210 seconds. As mentioned, one of the advantages of the present work is a
TE

more accurate prediction of CNG mass flow rate than the previous works.
C EP
AC

15
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
Fig. 9 Samples of mass flow rate of the gas for cascade storage system
M

The final results of the CNG refueling station are shown in Table 3. It should be noted that 26 vehicles could be
D

refueled in a cycle for the cascade storage system. The number of cycles per day is equal to 12. Thus, the total
numbers of on and off the compressor are almost 12. This is an important parameter because by increasing this
number, the compressor maintenance costs increase. The energy consumed by the compressor for a cycle,
TE

determined by using of the previous section (Theory and modeling) is equal to 61.86 kWh. The average specific
energy consumption is equal to 0.250 kWh/kg for a vehicle. Bang, H.J., et al. [8] has calculated average specific
energy consumption to be around 0.19 kWh/kg for slow filling process (Residential Refueling System). As the
energy consumption for slow filling is expected to be less than fast filling, the calculated value for specific energy
EP

consumption could be considred valid. This is an important value as it could give a rough estimation for calculating
energy consumption by the compressor in a CNG station. The value could be also used to design a more optimized
CNG refueling station.
C

Table 3. The final results of the cascade storage systems

AC

Required time for a cycle 122 min.

Number of vehicles that could be fueled in a cycle 26

Number of cycles per day 12

Number of vehicles that could be fueled in a day 307

Energy consumption for a cycle 61.86 kWh

Average filling time for a vehicle 242 sec.

16
 ACCEPTED MANUSCRIPT

Average specific energy consumption for a vehicle 0.250 kWh/kg

The best pattern for a system of storage reservoir could be selected by balancing the two parameters, filling time
and energy consumption by the compressor. This could be the subject of future works.

PT
4. Conclusion

RI
The review of literature shows there isn’t any research about of a CNG and CHG refueling station which
considers the reservoir tanks, on-board NGV’s cylinder, connection pipes and reciprocating compressor. It should be

SC
noted that in all previous work, only the fast filling process is modeled. The fast filling process is a part of a CNG
refueling station. The compressors have not been considered in all previous studies which investigated the CNG
stations. The main objective of the present work is to apply a thermodynamic analysis for a CNG refueling station,
including on-board NGV cylinder, reservoir tanks, connecting pipes between them and reciprocating compressor.

U
The results of the present work have been validated against the experimental data and compared with the results of a
previous work to investigate the effect of connecting pipes. On the other hand, unlike all previous research, pressure
AN
changes of the reservoir tanks during the filling process have been considered in the present work. The reciprocating
compressor of CNG refueling station has been modeled in the present work. The experimental data, measured by the
present authors, is used to identify the compressor performance in unsteady conditions. The modeling of connecting
pipes at present work cause the better dynamic pressure profile than the previous works. It is concluded that
M

variations in the compressor discharge pressure have no effect on the first and intermediate compressor pressure. It
has also no effect on the compressor delivery mass flow rate. It is found that 26 vehicles could be refueled in a cycle
for a cascade storage system. The energy consumed by the compressor for a cycle is equal to 61.86 kWh. The
D

average specific energy consumption is equal to 0.250 kWh/kg for a vehicle. The best pattern for a system of storage
reservoir could be selected by balancing the two parameters, filling time and energy consumption by the
TE

compressor. This could be the subject of future works.

5. Acknowledgments
EP

This work was supported by the research fund of National Iranian Oil Products Distribution Company
(NIOPDC). The authors thank for kind helps of the supporters.
C

Nomenclature
AC

A area, (m2)
cp Constant pressure specific heats (kJ/kg K)
cv Constant volume specific heats (kJ/kg K)
D Diameter (mm)
f Fanning friction factor
g Gravitational acceleration, (m/s2)
h enthalpy, (kJ/kg)

17
 ACCEPTED MANUSCRIPT

k Isentropic expansion factor

kWh kilowatt hour
l Liter
L Length (m)

 Mass flow rate, (kg/s)

PT
M Molecular weight, (kg/kmol)
P Pressure, (bar or kPa)

RI
Heat transfer rate, (kW)
R Gas constant (J/mol K)

SC
T Temperature, (K or oC)
u Internal energy, (kJ/kg)
t time, (s)

U
V Velocity (m/s)
AN
V Volume, (m3)
W Actual work, (kJ/kg)

 Actual work rate, (kW)

M

z Height, (m)

4 Density, (kg/m3)
D

Subscript
TE

a actual
avg average
c on-board NGV cylinder
co compressor
EP

cv control volume
i inlet condition
dis discharge
r reservoir tank
C

su suction
std standard
AC

References

1. Kountz, K., Modeling The Fast Fill Process in Natural Gas Vehicle Storage Cylinders. American Chemical
Society Paper at 207th National ACS Meeting, 1994.

2. Kountz, et al., NGV Fuelling Station and Dispenser Control Systems. report GRI-97/0398, Gas Research
Institute, Chicago, Illinois, 1997.

3. Kountz, K., W. Liss, and C. Blazek, Method and Apparatus For Dispensing Compressed Natural Gas. U.S.
Patent 5,752,552, 1998.

18
 ACCEPTED MANUSCRIPT

4. Kountz, K., W. Liss, and C. Blazek, Automated Process and System For Dispensing Compressed Natural
Gas. U.S. Patent 5,810,058, 1998.

5. Kountz, K., W. Liss, and C. Blazek, A New Natural Gas Dispenser Control System. International Gas,
Research Conference, San Diego, 1998.

6. Farzaneh-Gord, M., Compressed natural gas Single reservoir fi lling process. Gas international
Engineering and Management, 2008. 48(6): p. 16-18.

PT
7. Farzaneh-Gord, M., S. Hashemi, and A. Farzaneh-Kord, Thermodynamics Analysis of Cascade Reserviors
Filling Process of Natural Gas Vehicle Cylinders. World Applied Sciences Journal, 2008. 5(2): p. 143-149.

RI
8. Bang, H.J., et al. Modeling and Analysis of a CNG Residential Refueling System. in ASME 2014 Dynamic
Systems and Control Conference. 2014. American Society of Mechanical Engineers.

SC
9. Farzaneh-Gord, M., H. Reza Rahbari, and M. Deymi-Dashtebayaz, Effects of Natural Gas Compositions on
CNG Fast Filling Process for Buffer Storage System. Oil Gas Sci. Technol. – Rev. IFP Energies nouvelles,
2014. 69(2): p. 319-330.

U
10. Deymi-Dashtebayaz, M., et al., The full simulation of rapid refueling of a Natural Gas Vehicle on-board
cylinder. Journal of Natural Gas Science and Engineering, 2014. 21(0): p. 1099-1106.
AN
11. Olmos, F. and V.I. Manousiouthakis, Gas tank fill-up in globally minimum time: Theory and application to
hydrogen. International Journal of Hydrogen Energy, 2014. 39(23): p. 12138-12157.
M

12. Hosseini, M., et al., Thermodynamic analysis of filling compressed gaseous hydrogen storage tanks.
International Journal of Hydrogen Energy, 2012. 37(6): p. 5063-5071.

13. Okawa, Y., et al., Flow rate measurement via isothermal discharge method for hydrogen. International
D

Journal of Hydrogen Energy, 2012. 37(24): p. 18882-18887.

TE

14. Bourgeois, T., et al., Evaluating the temperature inside a tank during a filling with highly-pressurized gas.
International Journal of Hydrogen Energy.

15. Monde, M., et al., Estimation of temperature change in practical hydrogen pressure tanks being filled at
EP

high pressures of 35 and 70 MPa. International Journal of Hydrogen Energy, 2012. 37(7): p. 5723-5734.

16. Wang, G., et al., Investigations of filling mass with the dependence of heat transfer during fast filling of
hydrogen cylinders. International Journal of Hydrogen Energy, 2014. 39(9): p. 4380-4388.
C

17. Zhao, Y., et al., Numerical study on fast filling of 70 MPa type III cylinder for hydrogen vehicle.
AC

International Journal of Hydrogen Energy, 2012. 37(22): p. 17517-17522.

18. Suryan, A., H.D. Kim, and T. Setoguchi, Three dimensional numerical computations on the fast filling of a
hydrogen tank under different conditions. International Journal of Hydrogen Energy, 2012. 37(9): p. 7600-
7611.

19. Zheng, J., et al., Development of high pressure gaseous hydrogen storage technologies. International
Journal of Hydrogen Energy, 2012. 37(1): p. 1048-1057.

20. Zheng, J., et al., Experimental and numerical study on temperature rise within a 70 MPa type III cylinder
during fast refueling. International Journal of Hydrogen Energy, 2013. 38(25): p. 10956-10962.

19
 ACCEPTED MANUSCRIPT

21. Deymi-Dashtebayaz, M., M.F. Gord, and H.R. Rahbari, Studying transmission of fuel storage bank to NGV
cylinder in CNG fast filling station. Journal of the Brazilian Society of Mechanical Sciences and
Engineering, 2012. 34: p. 429-435.

22. Striednig, M., et al., Thermodynamic real gas analysis of a tank filling process. International Journal of
Hydrogen Energy, 2014. 39(16): p. 8495-8509.

23. Khadem, J., M. Saadat-Targhi, and M. Farzaneh-Gord, Mathematical modeling of fast filling process at

PT
CNG refueling stations considering connecting pipes. Journal of Natural Gas Science and Engineering,
2015. 26(0): p. 176-184.

RI
24. A Practical Guide to Compressor Technology2006: Wiley.

25. Stouffs, P., M. Tazerout, and P. Wauters, Thermodynamic analysis of reciprocating compressors.
International Journal of Thermal Sciences, 2001. 40(1): p. 52-66.

SC
26. Sun, S.-Y. and T.-R. Ren, New method of thermodynamic computation for a reciprocating compressor:
Computer simulation of working process. International Journal of Mechanical Sciences, 1995. 37(4): p.
343-353.

27.

U
Elhaj, M., et al., Numerical simulation and experimental study of a two-stage reciprocating compressor for
AN
condition monitoring. Mechanical Systems and Signal Processing, 2008. 22(2): p. 374-389.

28. Ndiaye, D. and M. Bernier, Dynamic model of a hermetic reciprocating compressor in on–off cycling
operation (Abbreviation: Compressor dynamic model). Applied Thermal Engineering, 2010. 30(8–9): p.
M

792-799.

29. Oosthuizen, P.H. and W.E. Carscallen, Compressible Fluid Flow. McGraw-Hill, 1997.
D

30. Weymouth, T.R., Problems in natural gas engineering. ASME, 1912. 34: p. 185-189.
TE

31. Vidal, J., Thermodynamics: Editions OPHRYS.

32. Menon, E.S., Transmission Pipeline Calculations and Simulations Manual2014: Elsevier Science.
EP

33. Hanlon, P., Compressor Handbook2001: McGraw-Hill Education.

34. George, D., CNG Sampling, in Natural gas sampling technology conference, K. Mayeaux, Editor 2014:
New Orleans, Louisiana. p. 72-80.
C
AC

20
 ACCEPTED MANUSCRIPT

• The main objective of the present work is to apply a thermodynamic analysis for a CNG
refueling station, including on-board NGV cylinder, reservoir tanks, connecting pipes and
reciprocating compressor.

• It could be realized that, the simulation result of mass flow rate in the present work is a

PT
good agreement with the experimental values.

RI
• The results show that variations in the compressor discharge pressure have no effect on
the first and intermediate compressor pressure.

U SC
AN
M
D
TE
C EP
AC