2964 Energy & Fuels 2007, 21, 2964-2968

Improved Designs of FCC Gasoline Hydrodesulfurization Units by

Properly Measuring the Olefin Content of the Gasoline Feed
Gary G. Podrebarac* and Arvids Judzis, Jr.
CDTECH, 10100 Bay Area BouleVard, Pasadena, Texas 77507

ReceiVed NoVember 21, 2006. ReVised Manuscript ReceiVed July 9, 2007

Hydrodesulfurization of cracked gasoline is now a vital step for the production of clean fuels. Along with
sulfur removal, however, olefin saturation occurs. Knowing the olefin content of the gasoline is key to achieving
a proper design, as olefin saturation largely sets the heat release and hydrogen consumption that will be
experienced. More specifically, the molar concentration of the olefins must be known, and determining this in
cracked gasoline is not as straightforward as it seems. There are a number of seemingly appropriate analytical
methods for olefin measurement. This study examines several of the more common methods used in the refining
industry and compares their performance on a sample of full-range FCC gasoline. A case is made that the
bromine number is the most appropriate measurement to use as the basis for a reactor design.

Introduction resolution GC, ASTM-D6733;2 (3) PIANO: detailed hydro-

The removal of sulfur from gasoline has become a vital
refining step as governments have enacted laws requiring cleaner
burning fuels. Of particular importance is the desulfurization
of cracked gasoline, which typically contributes the bulk of the
sulfur in the gasoline pool. In the capacity of a process licensor,
CDTECH has gained considerable experience in the area of
gasoline hydrotreating. CDTECH’s first commercial CDHDS
unit to desulfurize a heavy FCC gasoline fraction came on-
stream in May 2000, and many more designs have followed.
Olefin saturation is a side reaction that occurs as cracked
gasoline undergoes hydrodesulfurization (HDS). It is an undesir-
able exothermic reaction that consumes hydrogen and also
causes reduction of the octane rating of the gasoline product.
In cracked gasoline, the sulfur content may be 1000 ppmw or
0.017 mol of S/L, and the concentration of olefins may be 3
mol of Cd/L. These are just typical figures, but they illustrate
how the olefin concentration can be 2 orders of magnitude
higher than the sulfur concentration.
To properly design a gasoline HDS reactor, the heat balance
and the hydrogen consumption rates need to be known with
confidence. Given the relative concentrations of olefins and
sulfur, it is clear that the olefin content of the gasoline has a
greater impact on these engineering considerations than the
sulfur content. Thus, it is very important to know the molar
olefin concentration of the gasoline to be treated.
A variety of techniques is available to determine the olefin
content of gasoline, with most reporting results in vol % (or wt
%) olefins in an attempt to be historically consistent with the
FIA method (ASTM-D1319). During the development and
commercialization of CDTECH’s FCC gasoline HDS processes,
the following techniques were commonly encountered: (1) Br
#: bromine number, ASTM-D1159;1 (2) PONA: 50 m high-
* Corresponding author. Fax: (281) 474-0660; e-mail: gary.podrebarac@
cdtech.com.
(1) ASTM-D1159-93. Standard Test Method for Bromine Numbers of
Petroleum Distillates and Commercial Aliphatic Olefins by Electrometric
Titration; ASTM International: West Conshohocken, PA.
carbon analysis, ASTM-D5134;3 and (4) MD-GC Reformulizer
method, ASTM-D6839.4
Each of these methods is a reproducible technique. However,
we observed that comparisons of these various methods tend
to provide inconsistent results. This becomes especially apparent
when one tries to convert to molar units of concentration. Studies
in the literature also compare some of the measurement
techniques and also confirm that differences may occur.
In 1990, Kosal et al.5 compared FIA with a capillary column
technique (similar to the PONA and PIANO methods) and a
multidimensional (MD) column technique (similar to the MD-
GC/Reformulizer) to identify paraffins, naphthenes, and aromat-
ics. The authors were somewhat critical of the FIA method,
citing long analysis times and difficulty reading the exact
boundary between hydrocarbon types. In general, it was reported
that the two GC methods agreed with FIA but that problems
with coelution of components were observed. The coelution
problem grew worse as the boiling point of the sample increased,
and the authors subsequently confined the testing to those
samples with an endpoint of less than 215 °C (419 °F). The
multicolumn method worked quite well, even with some heavy
samples beyond the 215 °C endpoint. However, the authors
cautioned that this technique is quite a bit more complex than
the capillary method, requiring significantly more maintenance
work. No information was gathered on olefin measurements.
In a study by Barman,6 a comparison was made between FIA,
Br #, and an MD-GC method ASTM-D5443, commonly called
PIONA. The PIONA technique showed good agreement with
(2) ASTM-D6733-01. Standard Test Method for Determination of
IndiVidual Components in Spark Ignition Engine Fuels by 50 Meter
Capillary High Resolution Gas Chromatography; ASTM International:
West Conshohocken, PA.
(3) ASTM-D5134-98. Standard Test Method for Detailed Analysis of
Petroleum Naphthas through n-Nonane by Capillary Gas Chromatography;
ASTM International: West Conshohocken, PA.
(4) ASTM-D6839-02. Standard Test Method for Hydrocarbon Types,
Oxygenated Compounds, and Benzene in Spark Ignition Engine Fuels by
Gas Chromatography; ASTM International: West Conshohocken, PA.
(5) Kosal, N.; Bhairi, A.; Ali, M. A. Determination of hydrocarbon types
in naphthas, gasolines, and kerosenes: A review and comparative study of
different analytical procedures. Fuel 1990, 69, 1012-1019.

10.1021/ef060593k CCC: $37.00 © 2007 American Chemical Society

Published on Web 08/29/2007
Designs of FCC Gasoline Hydrodesulfurization Units
Table 1. Summary of Analytical Techniques To Be Compared
method description
Br # known weight of sample is dissolved
D1159 in a solvent, then titrated
with a bromide/bromate solution
electrometric detection of titration endpoint
PONA determines individual hydrocarbon components
D6733 of spark ignition engine fuels with boiling
ranges up to 225 °C (437 °F)
technique can be used for iindividual blend-stocks,
but statistical data for the method were only
gathered with blended fuels
PIANO determines individual hydrocarbon components
D5134 of petroleum samples
sample should have an endpoint by D-3710
less than 250 °C (482 °F)
MD-GC determines hydrocarbon types, including oxygenates
D6839 MTBE,TAME, ETBE, and ethanol
suitable for use on spark ignition engine fuels
the FIA results, but there was some scatter when correlating
the Br # technique against PIONA. Barman suspected that this
scatter was due to interference from sulfur compounds
and higher dienes. However, some results from CDTECH’s
internal tests indicated that sulfur compounds in FCC gasoline
are not a very significant source of error. The diene content of
FCC gasoline also tends to be fairly low relative to the olefin
content.
A study by Lubeck and Cook7 compared the results of FIA
with the Br # and found a reasonable agreement between the
techniques. However, they reported difficulties with the FIA
tests. They found an unusual nonlinearity in the FIA results as
the gasoline sample was diluted with an inert hydrocarbon. They
reported a significant error with FIA when samples were not
depentanized and were critical of the difficulty in determining
the position of the boundary between different hydro-
carbon types. It was also reported that the FIA technique
usually measured lower olefin concentrations than the Br #
technique.
Lubeck and Cook7 also discussed how the conversion from
Br # units to vol % olefins units can affect the results. A
conversion calculation is provided in the ASTM-1159 method,
but this calculation may not be adequate in every situation.
Perhaps Barman’s6 problems correlating Br # to PIONA were
due to samples that did not conform to the requirements of the
ASTM conversion correlation.
A study was conducted by Beens et al.8 in which three of
the 10 gasoline characterization techniques that were studied
were capable of olefin measurement: Reformulizer (ASTM-
D6839), FIA, and PONA (ASTM-D6733). Their goal was to
(6) Barman, B. N. Determination of hydrocarbon types in gasoline range
samples by multidimensional gas chromatography, fluorescent indicator
adsorption, and bromine number methods. Fuel 1995, 74, 401-406.
(7) Lubeck, A. J.; Cook, R. D. Bromine number should replace FIA in
gasoline olefins testing. Oil Gas J. 1992, 90, 86-89.
(8) Beens, J.; et al. A comparison of ten different methods for the analysis
of saturates, olefins, benzene, total aromatics, and oxygenates in finished
gasolines. J. Chromatogr. Sci. 2003, 41, 564-569.
Energy & Fuels, Vol. 21, No. 5, 2007 2965
limitations
samples need to be free of material lighter than
isobutane
samples should have D-86 90% distillation temperature
under 327 °C (626 °F)
this technique does not identify individual components
some minor variation in the response of different
olefins, which is quantified in Table A1.1 of the method
some sulfur compounds interfere with the titration results
generally, only one of the double bonds in a diene molecule is
titrated
benzene coelutes with 1-methyl-cyclopentene
toluene coelutes with 2,3,3-tri-methylpentane
when technique is used for PONA determination, user must be
cautioned about errors from coelution of components
procedure is applicable to samples with less than 20 mass %
olefins
significant coelution is possible with olefins >C7 and particularly
when samples are derived from FCC gasoline; total olefin
content with these samples may not be accurate
method is applicable to olefin-free liquid hydrocarbons
olefin content should be <2 vol % as determined by FIA
peaks eluting after n-nonane are generally reported as C10+
some coelution of compounds is acknowledged
cannot identify individual components, except for benzene
some overlap between C8/C9 aromatics
olefin content up to 30% v/v
oxygen content up to 15% v/v
aromatics content up to 50% v/v
find a method that gives essentially the same result as FIA but
with better precision. The data showed that all three olefin
methods provided similar average results but that the precisions
of the Reformulizer and PONA methods were much better than
FIA. The conclusion of the study was to recommend the
Reformulizer method as the preferred approach.
A summary of the relevant olefin analysis techniques is
presented in Table 1. This is not a complete listing but simply
a list of some techniques that are commonly used by refiners
to analyze their gasoline samples. FIA is not included in Table
1. Although FIA is very commonly referenced as a standard by
which other techniques are compared, it is a cumbersome
technique that many experts regard as being obsolete.
When one considers the limitations and capabilities of the
techniques in Table 1, it would appear that the Br # method is
likely the most suitable measure of aliphatic unsaturation in FCC
gasoline samples. There is, admittedly, some interference from
S compounds and dienes. However, the most significant S
compounds, such as thiophenes1 and benzothiophene, show
negligible responses to the Br # method. It is only compounds
such as mercaptans and sulfides that have the potential to impact
the measurement, and these components are present in very
small amounts. The amount of diene is also quite low as
compared to the olefin concentration, perhaps only 2 wt %. The
stated inaccuracies involved with the PONA and PIANO
methods seem to be much more substantial. Even with the
Reformulizer method, there is a valid concern that the olefin
content in certain fractions of FCC gasoline could exceed the
stated 30% limit.
From our viewpoint, one of the advantages of the Br # is
that it is directly related to the number of moles of olefin in the
sample. The determination of a molar concentration of olefin
is one of the goals of this paper because this is easily related to
hydrogen consumption and heat release when conducting a
reactor design.
This study examines how the results of three common olefin
measurement techniques compare when analyzing a sample of
2966 Energy & Fuels, Vol. 21, No. 5, 2007 Podrebarac and Judzis

Table 2. Properties of RFCC Gasoline Sample Table 3. PIANO Totals (Mass %) in RFCC Gasoline Sample
paraffins i-paraffins olefins naphthenes aromatics total
Method
C1 0.000 0.000 0.000 0.000 0.000 0.000
boiling range C2 0.000 0.000 0.000 0.000 0.000 0.000
ASTM-D3710 C3 0.000 0.000 0.000 0.000 0.000 0.000
ibp (°F) 42 C4 0.325 0.172 1.968 0.000 0.000 2.465
5% 94 C5 0.922 6.173 9.600 0.179 0.000 16.874
10% 101 C6 0.617 5.808 7.684 1.986 0.996 17.090
20% 144 C7 0.712 3.950 3.831 3.363 4.014 15.871
30% 170 C8 0.475 3.668 1.229 3.159 7.503 16.034
40% 199 C9 0.219 3.369 0.417 2.067 6.735 12.807
50% 231 C10 0.189 1.313 0.036 0.350 7.018 8.906
60% 261 C11 0.096 0.608 0.075 0.088 2.099 2.965
70% 290 C12 0.121 0.778 0.037 0.000 1.123 2.059
80% 339 C13 0.085 0.070 0.015 0.000 0.000 0.170
90% 396 total 3.762 25.908 24.891 11.193 29.487 95.242
95% 428 oxygenates 0.000 total C14+ 1.165
total unknowns 3.594 grand total 100.000
Ep 489
total S (ppmw) 740
density (g/mL) 0.7497 ASTM-D1159 method. Indeed, all three of these techniques
Br # (g/100 g) 63.2 seem to provide fairly consistent results when analyzing for wt
calcd olefins (wt %) by 28.85 % olefin.
A2.2 procedure
However, this is not the case when one converts the units of
PONA the PONA and PIANO measurements to match the units of the
av Mw (g/mol) 97.13
olefins (wt %) 28.77
Br #. The conversion is performed as follows:
calcd Br # (g/100 g) 47.3
olefin wt % × 159.8
PIANO calcd Br )
av Mw (g/mol) 92.46 Mw
olefins (wt %) 24.89
calcd Br # (g/100 g) 42.9 To use this equation, the engineer must have a good value
for the molecular weight (Mw). In many cases that we have
full-range cracked gasoline. There is not very much information encountered, the refiner provides an abbreviated output of the
in the literature with regard to neat samples of cracked gasoline. PONA or PIANO method, which only shows the wt % of olefin,
Only in the work of Lubeck and Cook7 were some neat FCC aromatics, etc. This leaves the engineer, who is not usually an
gasoline samples tested. This study also compares these analyti- expert in analytical chemistry, to make assumptions to fill in
cal techniques in different boiling ranges of the FCC gasoline. the Mw term. Frequently, as a first pass, this is assumed to be
the same as the average Mw of the gasoline sample. Results are
Experimental Procedures shown in Table 2 for calculations followed through on this basis.
It should be very clear that the calculated Br # values from the
A sample of cracked gasoline from a resid-FCC was obtained PONA and PIANO methods are substantially less than the
from a North American refinery for use in this study. This sample measured Br #.
was subdivided and submitted for Br #, PIANO, and PONA
As discussed, a common source of error in the conversion
analysis. An attempt was made to run the Reformulizer MD-GC
method, but no facility in the Houston, TX area was able to from wt % to Br # is the Mw term. The olefins in FCC gasoline
accommodate our request. are distributed preferentially to the light end. The average Mw
To measure the olefin concentration profile as a function of of the sample is higher than the average Mw of the olefin
boiling point, a 1 L sample of gasoline was batch distilled into 10 components. Thus, simply using the average molecular weight
× 100 cm3 fractions. The quantity of each gasoline cut was of the sample to convert the olefin content to a molar quantity
sufficient to run the various analytical tests, such as total S, Br #, is a very poor assumption. The design engineer should ensure
density, PONA, etc. that the full output from the PONA or PIANO method is
The fractionation was conducted in a 2 in. diameter × 36 in. provided. For example, a PIANO output of the components
high Oldershaw column. The column was vacuum jacketed for sorted by carbon number is shown in Table 3. On the basis of
insulation and contained 10 distillation trays. The condenser was the distribution of olefins provided in Table 3, the average Mw
cooled with a circulating cooling bath. The first eight cuts were for olefin components is 78.4 g/mol for the PIANO method,
obtained at atmospheric pressure, with a reflux ratio of 5:1. For
and a similar analysis works out to 81.46 g/mol for the PONA
the ninth cut, the reboiler temperatures became too high, so the
column was placed under vacuum to permit operation at lower method. Using these revised average Mw values for the olefin
temperatures. The ninth cut was collected overhead under vacuum, gives a calculated Br # of 50.5 g/100 g for the PIANO method
and the final tenth cut consisted of the material remaining in the and 56.4 g/100 g for the PONA method. Even with the corrected
reboiler. Mw, the Br # calculated from the PIANO analysis is still far
from the measured Br #. The measured Br # of 63.2 is 12%
Results and Discussion higher than the Br # based on PONA.
Curiously, the ASTM conversion of Br # to wt % olefins
The properties of the FCC gasoline sample are provided in gives almost perfect agreement, but the conversion of PONA
Table 2. Clearly, there are differences between the wt % olefin olefin to a Br # gives about 12% error. The difference seems to
by the PONA method and the wt % olefin by the PIANO reside within the correction factor published in section A2.2,
method. However, this magnitude of disagreement may not be Table A2.1 of ASTM-D11591. This correction factor compen-
too alarming. There is excellent agreement between the PONA sates for the boiling range of the sample and has been fitted to
method and the Br # method when the Br # is converted to wt allow the Br # to agree with PONA and/or FIA data. Conversely,
% olefins using the procedure outlined in section A2.2 of the when converting the units of the PONA and PIANO methods
Designs of FCC Gasoline Hydrodesulfurization Units
Figure 1. Comparison of Br #, PONA, and PIANO.
to match the Br # unit, no correction factor is required. Yet,
the results differ.
It is interesting that both GC-based techniques provide a lower
olefin concentration measurement than the Br # method. Lubeck
and Cook7 reported a similar observation when comparing the
Br # and FIA. One cause is the algorithm and/or chemistry used
to group the components. Generally, components like cyclo-
hexene are classed as olefins, not naphthenes. However, an
alkenyl aromatic component like styrene (Mw ) 104), which
consists of an olefinic side group attached to an aromatic ring,
is classified as an aromatic. These olefinic functionalities are
reactive under HDS conditions and are detected by the Br #
but are not reported as olefins by the PIANO or PONA methods.
Another problem with the PONA and PIANO methods is
coeluting components. Table 1 highlights the coelution problem
as being very significant for the PONA method with C7 and
heavier components. An analysis of the series of 10 narrow
gasoline cuts can provide some interesting insight into the
behavior of the PONA and PIANO methods with various boiling
ranges of cracked gasoline.
Figure 1 is a plot of the Mw and olefin content of each cut as
a function of the 50% point of the cut. The Mw was taken from
the PONA analysis of each individual cut. The Mw from the
PIANO analysis was not used because there were too many
unknown components (about 18 wt %) to make an accurate
determination of the Mw in the heaviest cut.
Figure 1 illustrates that there is reasonable agreement between
Br #, PONA, and PIANO techniques with the light cuts of FCC
gasoline. However, as the boiling point of the cuts exceeds 200
°F, a significant divergence of the three methods is observed.
Above 200 °F, the Br # technique always measures the highest
olefin concentration, with the PONA measurement below that,
and the PIANO measurement measuring the lowest olefin
concentration.
Figure 2 shows the relative difference observed between the
PONA and PIANO methods when compared to the Br #. The
differences are initially about 10%. However, as the Mw
approaches 100 g/mol, the gap widens. Near 100 g/mol, there
is a very steep increase in the difference. It is probably not
coincidental that the molecular weight of a C7 olefin is 98.2
g/mol and that the molecular weight of styrene is 104. It appears
that coelution of components, which was highlighted in Table
1, could very well cause a loss of accuracy in the PONA and
PIANO techniques. In addition, the classification of alkenyl
aromatics as aromatics would also start to cause differences at
this point in the boiling range.
As the FCC gasoline becomes heavier, the fractions also tend
to become increasingly aromatic in nature. This is also shown
Energy & Fuels, Vol. 21, No. 5, 2007 2967
Figure 2. PONA and PIANO relative to Br #.
Table 4. Example of Treating a 160-450 °F FCC Gasoline
Feedstock for HDS
Br # PONA PIANO
feed total S (ppmw) 940 940 940
feed density (g/mL) 0.79 0.79 0.79
feed Br # (g/100 g) 52.5 33.6a 23.7a
H2 consumption (scf/bbl) 141.7 90.7 63.9
heat release (Btu/bbl) 18804 12028 8477
a Calculated Br # starting from either PONA or PIANO method.
in Figure 2. The Br # continues to detect the presence of olefin
compounds in the heavy end of the FCC gasoline, with a value
of about 20 g of Br2/100 g in the heaviest cut. However, the
PONA and PIANO methods show an olefin content of es-
sentially zero, or about 100% error. It was mentioned earlier
that most of the olefins present in this fraction of the gasoline
are likely in the form of alkenyl aromatics. As such, these would
be classified as aromatics by the PIANO and PONA algorithms.
Clearly, PONA and PIANO give the reactor design engineer a
misleading impression that there is no aliphatic unsaturation in
this portion of the gasoline.
In most FCC gasoline desulfurization designs, the light
gasoline is fractionated away from the heavy gasoline and treated
separately. This is done to preclude hydrogenation of the highly
concentrated light olefin in the HDS reactor. This results in a
boiling range of roughly 160-450 °F being treated in the HDS
reactor, which can be approximated by mathematically adding
the heaviest seven cuts from Figures 1 and 2. As an exercise,
the performance of the reactor was calculated to see the
differences involved when using Br #, PONA, or PIANO as
the analytical method of choice.
Table 4 is a summary of the heat release and hydrogen
consumption that would be predicted. The basis for this
calculation is a 95% conversion of the sulfur in a fixed bed
reactor with performance characteristics published by Rosvoll
et al.9 As shown, there is a drastic difference in the predicted
hydrogen consumption and heat release. Using the PONA
method, the heat release is understated by 56%. Using the
PIANO method, the heat release is understated by 121%.
Clearly, a danger is that the refiner may not have enough
hydrogen available to run the reactor at full rates if the reactor
predictions were based on PONA or PIANO methods. Another
constraint is caused by heat integration in the feed/effluent heat
exchanger that results in the reactor not being able to process
the full feed rate for which it was intended.
(9) Rosvoll, J. S.; Grande, K.; Molijord, K.; Skjølsvik, P. A.; Thorvaldsen,
B.; Johansen, S.; Kuhnle, S.; Sørum, P. Å.; Sapre, A. V.; Roundtree, E. M.
Successful Start-Up of New SCANfining Unit at Statoil’s Mongstad Refinery,
8th Annual European Refining Technology Conference, London, 2003.
2968 Energy & Fuels, Vol. 21, No. 5, 2007
Octane loss may also be understated if olefin conversion is
not based on the Br #. For example, if the previous feedstock
is desulfurized (e.g., 95% sulfur conversion) in a fixed bed
hydrotreater, an octane loss prediction based on olefin measure-
ments by the PIANO method would be understated by 35%
relative to the octane loss prediction based on the Br #
conversion. This can have severe economic implications for the
refiner.
Conclusion
Three methods for measuring olefin concentration in FCC
gasoline have been compared. Two GC-based methods, PONA
and PIANO, were compared to the results of a Br # titration.
The ASTM method descriptions of PONA and PIANO clearly
state the limitations of using these methods to characterize
olefins. The results of this study confirm that these limitations
are generally valid. However, despite the stated limitations, these
methods find common applications in the refining industry in
areas where their accuracy is questionable.
An analysis of a series of narrow cuts of the FCC gasoline
shows that these analytical techniques are in good agreement
for light gasoline fractions. However, for gasoline fractions
containing C7 and heavier olefins that boil greater than 200 °F,
the PONA and PIANO methods diverge significantly from the
results of the Br #. In the heaviest fraction of the gasoline, the
PONA and PIANO methods were in error by essentially 100%.
A key problem is that they both classify alkenyl aromatic
compounds, such as styrene, as aromatics. While one can make
a case that this is the proper classification of these compounds
based on chemistry, this classification does not help the reactor
designer.
When designing HDS reactors, the saturation of olefinic side
groups releases heat, consumes hydrogen, and therefore must
be quantified. Design engineers should take care to fully
Podrebarac and Judzis
understand the definition of olefin used by the PONA and
PIANO methods and take the time to request and review the
full output from the PONA or PIANO methods. Regardless of
this effort, however, the designer still needs some measure of
aliphatic unsaturation that will aid in calculating the hydrogen
uptake and heat release in the reactor when dealing with heavy
gasoline. It should be clear that the Br # characterizes all of the
reactive olefins in the cracked gasoline and helps prevent design
errors.
At this point, the danger of designing an HDS reactor for
cracked gasoline based on PONA or PIANO data should be
clear. Both of these techniques tend to understate the amount
of olefin in the feed. If this is carried through the design
calculations, the heat release, hydrogen consumption, and octane
losses will be underestimated. Once the reactor is built, the
refiner will experience problems reaching the desired sulfur
conversion and problems with feed/effluent heat integration. The
reactor unit could also be constrained by insufficient hydrogen
availability. Ultimately, these complications could restrict overall
unit throughput.
It is also demonstrated that the deviation in the olefin
measurements between the PONA and the PIANO techniques
and the Br # method become more pronounced with heavier
fractions. This study was conducted on a full-range sample,
which includes C5 material. In practice, it is quite common to
remove the C5 and C6 fractions of the gasoline and treat the
C7+ gasoline fraction in the HDS reactor. This is precisely the
point where the PONA and PIANO techniques project a lower
olefin content and heat release. Thus, designs to treat heavy
FCC gasoline based on PONA or PIANO data could be
substantially undersized and/or unable to cope with the heat
release in this situation.
EF060593K