Validation of Diesel Fraction Content in Heavy Oils Measured by High Temperature Simulated Distillation and Physical Vacuum Distillation by Performance of Commercial Distillation Test and Process Simulation

Abstract

A gas chromatography high temperature simulation distillation (HTSD: ASTM D 7169), and physical vacuum distillation (ASTM D 1160) were employed to characterize H-Oil vacuum distillates, straight run vacuum distillates, and hydrotreated vacuum distillates with the aim to determine their content of diesel fraction and evaluate the possible higher extraction of diesel fraction from the heavy oils. The ASTM D 7169 reported about six times as high diesel fraction content in H-Oil heavy distillates as that reported by the ASTM D 1160 method. Performing a commercial distillation column test along with a simulation of the column operation using data of both ASTM methods and a software process simulator revealed that the HTSD is the more valid method for proper determination of the diesel fraction content in heavy oils. The software process simulation of the commercial distillation column operation suggests that the HTSD could be considered as a true boiling point distillation method for heavy oils. The separation of the diesel fraction from the H-Oil heavy distillates quantified by the HTSD could deliver oil refining profit improvement in the amount of six digits USD per year.

1. Introduction

Petroleum refining economics favor increased production of light oil products (boiling below 360 °C) at the expense of decreased production of heavy oils (boiling above 360 °C) [1,2]. The yields of light oil products depend not only on the characteristics of the crude oil processed and the conversion level in the heavy oil conversion units, such as fluid catalytic cracking, hydrocracking, and thermal cracking (visbreaking, or coking), but also on the efficiency of the distillation columns separating the light oils from the heavy oils [3,4,5,6]. The efficiency of the distillation column performance is typically controlled by examination of the amount of light oil content remaining in the heavy oil and the amount of heavy oil remaining in the light oil [7]. This is carried out by laboratory distillation tests following the ASTM standards of the distillation column products. Over the years, gas chromatography (GC) simulated distillation methods have found widespread application in the refining industry to help control the quality of fractionation taking place in commercial oil refining distillation columns, especially in those which fractionate mixtures containing heavy oils [8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23]. Because of their greater precision, lower cost per sample, reduced manual work, and safer process, simulated distillation methods replaced the physical vacuum distillation methods ASTM D 1160 and ASTM D 5236 [20]. Nevertheless, the suppliers of simulated distillation gas chromatography equipment provide correlations to convert the simulated distillation data into the physical distillation methods ASTM D 86 and ASTM D 1160 for more convenient use by the refiners. While the GC simulated distillation ASTM D 2887 was proved to be equivalent to the true boiling point (TBP) distillation according to the standard ASTM D 2892 [24] for light oils, the high temperature simulated distillation HTSD (ASTM D 7169) has not yet proved its equivalency to the TBP distillation for heavy oils. The different boiling point distributions reported by the HTSD and the ASTM D 1160 raise the following question: which distillation method provides more reliable and valid information about the remaining amount of the more valuable light oil in the less valuable heavy oil? Figure 1 illustrates the lower yield of diesel fraction in samples of heavy oils (atmospheric residue, straight run vacuum distillate, hydrotreated vacuum distillate, H-Oil heavy atmospheric gas oil (HAGO), light vacuum gas oil (LVGO), heavy vacuum gas oil (HVGO), and fluid catalytic cracking (FCC) feed hydrotreater diesel fraction when the ASTM D 1160 method is used.

Understandably, the question arises whether it is possible to extract more diesel from the heavy oils. If one uses the ASTM D 1160 distillation data, the answer to this question could be no, but if the ASTM D 7169 is used, the answer could be yes. Another important question, which is also related to the evaluation of the feasibility to extract more diesel fraction from the heavy oil products, is which distillation data should be used in the process of the commercial distillation column performance simulation? Kaes [25], for example, recommends the use of the ASTM D 1160 standard to characterize heavy oils instead of the simulation distillation in process simulation. While, Golden et al. [26,27,28] strongly recommend the employment of ASTM D 7169 when commercial distillation column operation simulation is concerned. A commercial distillation test at an operating distillation column separating mixtures containing heavy oils and simulation of its performance by a software process simulator may give an answer which laboratory distillation method ASTM D 7169 or ASTM D 1160 reports the correct value for the heavy oil diesel fraction content. For that reason, a commercial test at a FCC feed hydrotreater main fractionator was organized and performed in which the heavy oil streams were characterized by ASTM D 7169 and ASTM D 1160 methods. A simulation of the FCC feed hydrotreater main fractionator operation was carried out using a software process simulator and both ASTM D 7169, andASTM D 1160 distillation data.

The aim of this research is to verify which method: the ASTM D 7169 or ASTM D 1160, provides more reliable and valid information about the distillation characteristics of heavy oils, which can be used to optimize the process of extraction of the more valuable diesel fraction from heavy oils.

2. Materials and Methods

2.1. Materials and Laboratory Tests

Heavy atmospheric gas oil (HAGO), light vacuum gas oil (LVGO), and heavy vacuum gas oil (HVGO) from a commercial H-Oil ebullated bed vacuum residue hydrocracker with property range of variation shown in our recent research [29] were subject to ASTM D 7169 and ASTM D 1160 distillation analysis. Besides, straight run vacuum distillate, stabilized hydrotreated vacuum distillate, and diesel fraction from a FCC feed hydrotreater were also subject to ASTM D 7169 [30] and ASTM D 1160 [31] distillation analysis. Density at 15 °C of the studied oils was measured in accordance to ASTM D 4052 method [32].

The physical vacuum distillation analyses were carried out in Euro Dist MPS (ROFA) according to ASTM D-1160 requirements. The pressure in the Euro Dist MPS ASTM D-1160 apparatus during the whole analysis was 1 mm Hg.

The HTSD analyses were carried out with the Agilent Technologies GC System 7890B, which was equipped with FID (flame ionization detector). Liquid nitrogen was used as a coolant. The carrier gas was helium with 99.9999% purity (14 mL/min), the inlet pressure was 1.2 psi (8.27 kPa) with the total flow equal to 87 mL/min. Hydrogen was used as a fuel gas (40 mL/min) and nitrogen was a makeup gas (15 mL/min), both with high purity (99.999%). The installed column was 5 m long, 530 μm in diameter and the film thickness was 0.15 μm. The oven operated under the program from −20 °C to 430 °C at a ramp rate of 15 °C/min and a 4 min hold time at the maximum temperature. The injector was programmed to operate from 50 °C to 450 °C, at a rate of 15 °C/min, injected sample volume was 4 μL. Before the simulated distillation analyses of the studied oil samples they were stirred preliminary, accurately weighted to achieve 2 weight percent of studied oils dissolved in carbon disulphide (0.03 g of the sample dissolved in 1.5 mL CS₂ (99.9%)). All prepared samples were stored at a temperature around 4 °C prior to analyses. The simulated distillation characteristics were automatically calculated by the SIMDIS software and the distillation curve boiling point in °C versus evaporate in wt.% was obtained. Minor intervention of the operator took place during the chromatograms processing. The HTSD GC was calibrated with a blend of normal paraffins having a carbon number between C₅ and C₁₂₀. The software (GC OpenLab CDS with Simdis program for ASTM D-7169) used in this application of HTSD allows estimating of the final boiling point of the residual oils higher than 750 °C.

The repeatability of both studied laboratory distillation methods ASTM D 7169 and ASTM D 1160 applied to the H-Oil heavy distillates investigated in this work are summarized in

Table 1. Repeatability of the distillation methods ASTM D 7169, and ASTM D 1160 applied to the studied H-Oil heavy distillates.

Evaporate, %	95% CL (ASTM D 7169), °C	95% CL (ASTM D 1160), °C
IBP	9	17
5	3.0	3.2
10	3.0	3.2
20	3.0	3.2
30	4.0	3.2
40	4.0	3.2
50	4.0	3.2
60	4.0	3.3
70	4.0	3.3
80	4.0	3.3
90	4.0	3.3
Cost per a single analysis, USD	104	64

. The data in Table 1 also include the costs of performance of a single analysis by both methods, expressed in USD.

2.2. Commercial Distillation Test

The commercial test was carried out at the LUKOIL Neftohim Burgas (LNB) fluid catalytic cracking (FCC) complex unit. It consists of feed hydrotreating, FCC reactor-regenerator and fractionation sections. A process diagram of the LNB FCC complex is presented in Figure 2. More details about the LNB FCC complex unit is presented in our earlier research [33].

The commercial test consisted in processing of a definite quantity of a blend from H-Oil heavy atmospheric gas oil, light, and heavy vacuum gas oils which were characterized by the ASTM D 1160 method, and the ASTM D 7169 method in the stabilization column of the FCC feed hydrotreating section (see Figure 2). The stabilization column of the FCC feed hydrotreating section has three beds filled with Sulzer structured packing and operates at an absolute pressure of 1.5 bars. It fractionates the reaction mixture leaving the FCCPT reactors into gas, naphtha, diesel, and stabilized hydrotreated vacuum gas oil (VGO). The diesel fraction is separated from the second packing bed and is stripped with steam in a stripping column as shown in Figure 2. Two cases were studied in this work. The first one was that of the typical FCCPT stabilization column operation. The second one was that of the typical FCCPT stabilization column operation plus processing of additional 24.4 wt.% of H-Oil vacuum gas oils (of fresh FCCPT feed).

Figure 3 presents a flow chart describing the methodology and procedures to validate the diesel fraction content in H-Oil VGO measured by ASTM D 7169 and ASTM D 1160

Legend explanation:

G = content of diesel fraction, measured by ASTM D 7169, wt.%;

E = content of diesel fraction, measured by ASTM D 1160, vol.%

X = FCCPT fractionation section feed rate, t/h;

Y = FCCPT fractionation section diesel product rate, t/h;

Y + ΔY = FCCPT fractionation section diesel product rate when H-Oil VGO is processed, t/h;

Z = H-Oil VGO processed in FCCPT fractionation section, t/h;

B = H-Oil diesel flow rate estimated as a multiplication of H-Oil VGO flow rate by the content of diesel fraction, measured by ASTM D 7169, t/h;

C = H-Oil diesel flow rate estimated as a multiplication of H-Oil VGO flow rate by the content of diesel fraction, measured by ASTM D 1160, multiplied by the density of H-Oil diesel, t/h;

Comparison of ΔY with B, and C

If ΔY ≈ B (C), then ASTM D 7169 (ASTM D 1160) correctly quantifies diesel content in H-Oil VGO;

If ΔY > B (C), then ASTM D 7169 (ASTM D 1160) underestimates the diesel content in H-Oil VGO;

If ΔY < B (C), then ASTM D 7169 (ASTM D 1160) overestimates the diesel content in H-Oil VGO.

2.3. Commercial Distillation Process Simulation

A software process simulator was employed to simulate the FCCPT main fractionator operation for both cases using as input process stream characteristics those obtained by the ASTM D 1160, and the (ASTM D 7169 methods. All distillation curves were converted to True Boiling Point (TBP) curves using published in literature methods [34,35]. ASTM D86 distillation curves were converted with API Procedure 3A1.1 to TBP Curves [34]. ASTM D1160 distillation data are entered as atmospheric pressure equivalent temperature curve (AET) and is converted to TBP using the procedure described in API Technical Data book [35]. The ASTM D 2887 (simulated distillation) curve is converted by a two-step process—first to weight % ASTM D86 curve and then to TBP in accordance with the procedure described by Riazi [34]. HTDS (ASTM D7169) distillation curves are entered as a weight % true boiling point curves for the lack of a correlation to convert it to TBP Curve. Standard methods were applied for the generation of the pseudo-components and their molecular weight, critical properties and physical properties: New API method (Riazi–Daubert extended method) for molecular weight prediction, API Method for critical properties prediction [34].

Based on the measured flow rate of all outlet streams of the FCCPT stabilization column, the characteristics of the column inlet stream were obtained through summation. A simulation model was created for the stabilization section where the stabilization column is located. The Grayson–Streed equation [36] was used for the phase equilibrium calculations and the Lee–Kesler equation [37,38] was used for the enthalpy calculations.

The simulation model was fine-tuned to match the instrument readings as close as possible. The instrument readings are from a snapshot of the FCCPT Stabilization Section actual operation at the moment the product samples were collected (usually at 8:00 am). The only variable that was left free was the flowrate of the diesel fraction from the side-stripper column K-102.

3. Results

3.1. Laboratory Test Results

Figure 4 summarizes the variation of diesel fraction content (boiling ≤ 360 °C) in H-Oil HAGO, and LVGO for 2021 and 2022, when ASTM D 7169 is used. It is evident from these data that diesel in HAGO varies between 10.4 and 53.4 wt.%, averaging at 21.7 wt.%. Whereas the diesel in LVGO varies between 18 and 42 wt.%, averaging at 24.8 wt.%. It is clear from these data that a sufficient amount of diesel fraction remains in the H-Oil heavy distillates, which is unfavorable from an oil refining economic point of viewm and presents a strong incentive to recover the diesel fraction from H-Oil HAGO, and LVGO. However, the physical distillation data obtained by the use of ASTM D 1160 exhibits much lower content of diesel fraction in H-Oil HAGO, and LVGO as shown in Figure 5. The data in Figure 5a indicate 2.3 vol.% diesel fraction content in HAGO, according to ASTM D 1160, versus 12.4 wt.% according to the ASTM D 7169 method. The data in Figure 5 b indicate 5.0 vol.% diesel fraction content in LVGO, according to ASTM D 1160, versus 18.2 wt.% according to the ASTM D 7169 method. The ASTM D 1160 results disprove the opinion of poor diesel extraction in the H-Oil fractionator columns based on the ASTM D 7169 results. In order to validate which method provides more reliable information about the content of diesel fraction in H-Oil heavy distillates, a decision was made to perform a commercial test in which the H-Oil heavy distillates were to be processed in the FCCPT main fractionator.

3.2. Commercial Distillation Test Results

Table 2. Operating conditions during the commercial test without (Case 1) and with (Case 2) processing of H-Oil heavy distillates in the FCC feed hydrotreater stabilization column.

Process Parameters in the LNB FCC Feed Hydrotreating Section	Unit	Case 1	Case 2
Charge heater inlet temperature	°C	339	340
Inlet pressure in the charge heater	bar	44	44
First hydrotreating reactor inlet temperature	°C	357	357
First hydrotreating reactor inlet pressure	bar (g)	41	41
Inlet temperature in the heater, heating the stabilization column feed	°C	301	302
Stabilization column feed inlet temperature	°C	374	375
Stabilization column top temperature	°C	100	100
Stabilization column first packing bed temperature	°C	182	175
Stabilization column second packing bed temperature	°C	293	298
Stabilization column bottom temperature	°C	359	355
Stabilization column top pressure	bar (g)	0.39	0.4
Stabilization column bottom pressure	bar (g)	0.52	0.52
Steam rate in the bottom of the stabilization column	kg/h	3852	3952
Steam rate in the stripping column	kg/h	120	120
Hydrotreated VGO (bottom) product temperature	°C	374	376
Fuel gas rate in the heater, heating the stabilization column feed	nm3/h	1562	1779
Stabilization column wash oil rate	m3/h	9.0	11.6
First pump around rate (below the first packing bed)	m3/h	127	125
Second pump around rate (below the second packing bed)	m3/h	30	31
First pass rate during the heater, heating the stabilization column feed	m3/h	72	69
Second pass rate during the heater, heating the stabilization column feed	m3/h	68	69
Third pass rate during the heater, heating the stabilization column feed	m3/h	83	83
Fourth pass rate during the heater, heating the stabilization column feed	m3/h	74	77
Fifth pass rate during the heater, heating the stabilization column feed	m3/h	131	126
Sixth pass rate during the heater, heating the stabilization column feed	m3/h	134	128
Feed flow rate of the FCC feed hydrotreating section	t/h	160	160
Flow rate of H-Oil heavy distillate blend	t/h	0	39
Percent of opening the valve providing recycling during the heater, heating the stabilization column feed	%	19	3

summarizes the operating conditions in the FCC feed hydrotreating section during the commercial test with and without processing H-Oil heavy distillates in the FCC feed hydrotreater stabilization column. For a better understanding of the meaning of the process parameters shown in Table 2, a more detailed diagram of the FCC feed hydrotreating section stabilization column and the stripping column is given in Figure 6.

It is evident from the data in Table 2 that the operating conditions in the FCC feed hydrotreating section were almost the same for both studied cases. This may imply that any additional quantity of produced diesel in the FCC feed hydrotreating section should come from the additional H-Oil VGO blend processed in the stabilization column.

Table 3. Quantity of feeds and products of the FCCPT (FCC feed pretreater (hydrotreating section) obtained during the commercial test without (Case 1) and with (Case 2) processing of H-Oil heavy distillates in the FCC feed hydrotreater stabilization column.

FCCPT Feed Component Flow Rate	Without H-Oil VGO Processing in the FCCPT Stabilization Column (Case 1)		With H-Oil VGO Processing in the FCCPT Stabilization Column (Case 2)		Δ Diesel Fraction, t/h	% of Δ Diesel Fraction of H-Oil VGO Processed in the FCCPT Stabilization Column
FCCPT VGO feed rate, t/h	160		160
H-Oil VGO entering in FCCPT stabilization column, t/h			39
Hydrogen make up, t/h	0.85		0.83
FCCPT Products	t/h	yields, wt.%	t/h	yields, wt.%
FCCPT gas flow rate, t/h	1.1	0.7	1.2	0.7
FCCPT H2S flow rate, t/h	3.8	2.4	3.7	2.3
FCCPT naphta flow rate, t/h	2.1	1.3	1.8	1.2
FCCPT diesel flow rate, t/h	16.6	10.4	20.0	12.5	3.4	8.7
Hydrotreated VGO flow rate, t/h	137.0	85.6	172.9
Losses, t/h	0.2	0.1	0.2

presents data of the material balance of the LNB FCC feed hydrotreating section (called also FCC pretreater) for both studied cases in this research: Case 1—without processing H-Oil heavy distillates (VGO = vacuum gas oil); and Case 2—with processing H-Oil VGO. The data in Table 3 indicates that the processing of 39 t/h H-Oil VGO blend resulted in 3.4 t/h higher amount of produced diesel fraction from the FCC feed hydrotreating section. This equals to 8.7 wt.% of the H-Oil VGO processed in the FCCPT stabilization column.

Table 4. Characteristics of FCCPT feed, HTVGO, and diesel, and those of the H-Oil heavy distillates measured by ASTM D 1160, and the ASTM D 7169 methods for Case 2.

Parameter	FCCPT Feed	FCCPT Diesel	FCCPT HTVGO	H-Oil VGO Blend	H-Oil HVGO	H-Oil LVGO	H-Oil HAGO
Density at 15 °C, g/cm3	0.9254	0.8895	0.9120	0.9392	0.9539	0.9490	0.9344
Sulphur, wt.%	2.08	0.105	0.207	0.61	0.79	0.63	0.54
Distillation, wt.%/°C	ASTM D 7169
0.5	349	213	348	320	397	317	313
5	367	240	365	347	416	341	342
10	381	260	377	361	429	354	355
30	418	303	411	397	463	390	386
50	446	330	439	428	489	419	410
70	485	352	474	466	515	452	438
90	535	397	525	524	553	515	495
95	561	434	551	556	577	543	531
99.5	638	521	615	633	655	593	625
Recovery, %	104.8	88.9	104.2	103.0	112.1	101.4	100.2
Boiling up to 360 °C, wt.%	3.3	76.4	3.7	9.6	0.0	13.0	12.8
Distillation, vol.%/°C	ASTM D 1160
0.05	373	225	376	343	426	271	324
5	394	282	394	374	456	366	374
10	404	297	404	386	463	377	382
30	432	330	426	411	483	405	403
50	459	345	450	434	495	425	416
55	467	349	458	442	500	428	422
60	475	353	465	449	506	438	427
65	482	358	471	459	511	447	434
70	490		478	466	518	456	441
80	505		495	483	529	476	456
85	515		506	494	536	487	466
90	525		519		543	496	477
Boiling up to 360 °C, vol.%	0	66.9	0	1.7	0.0	4.5	1.1

presents data for the characteristics of FCCPT feed, HTVGO, and diesel, and those of the H-Oil heavy distillates measured by the ASTM D 1160 and the ASTM D 7169 methods for Case 2. It is worth mentioning here that the characteristics of the FCCPT feed, HTVGO, and diesel for Case 1 were identical with those of Case 2. The data in Table 4 indicates again a very big difference in the content of diesel fraction (boiling ≤ 360 °C) in the H-Oil VGO blend, that is additionally processed in the FCCPT stabilization column, measured by ASTM D 7169 (9.6 wt.%) and by the ASTM D 1160 (1.7 vol.%). If one takes a look at the data in Table 3, it can be seen that the additional amount of diesel separated from the FCCPT stabilization column, that comes from the H-Oil VGO blend, presents 8.7 wt.% of the H-Oil VGO processed in the column. This value is very close to that measured by the ASTM D 7169 and too far from the value reported by the ASTM D 1160 method. It should be pointed out here that the characteristics of the diesel fraction separated from the FCCPT stabilization column and the following stripping column were the same for both studied cases. This means that the increased quantity of produced diesel fraction from the FCCPT stabilization column and the following stripping column comes from the increased content of diesel fraction in the H-Oil VGO blend.

Table 5. FCC section operating conditions and FCC yields registered for Case 1, and Case 2 studied in this work.

Parameter	Case 1	Case 2
	without H-Oil Diesel Separation	with H-Oil Diesel Separation
FCC feed rate, t/h	174	173
H-Oil VGO in FCC feed, wt.%	21.8	20.6
H-Oil diesel in FCC feed, wt.%	2.0	0.0
Riser outlet temperature, °C	550	550
Combined feed temperature, °C	313	318
Regenerator dense bed temperature, °C	704	703
Regenerator dilute bed temperature, °C	692	691
Air rate for coke burning, kNm3/h	105.3	106.3
CTO ratio	9.59	9.68
Heat of reaction, kJ/kg	599	617
Delta coke, wt.%	0.54	0.54
Yields and conversion, wt.%
Dry gas	5.69	5.82
C3 fraction	8.59	8.62
C4 fraction	14.00	14.16
Gasoline	43.90	44.14
LCO	7.71	7.01
HCO	8.81	9.36
Slurry oil	6.05	5.68
Coke	5.22	5.21
Losses	0.03	0.01
Total	100.00	100.00
Conversion	77.43	77.96

presents data of FCC section operating conditions and FCC yields registered for Case 1, and Case 2 studied in this work. This data shows that the separation of diesel fraction from the H-Oil VGO not only increases the amount of diesel produced in the refinery, but also enhances the conversion in the FCC unit, at the expense of lowering the light cycle oil (LCO) yield. By doing an economical evaluation using the LP (linear programming) model of the LNB refinery, the margin improvement from the separation of diesel fraction from the H-Oil VGO amounts to about 7 MM USD/year.

3.3. Distillation Process Simulation Results

The results from the simulation of the FCCPT stabilization column operation by the use of the software process simulator are presented in the Supplementary Materials as Figures S1–S6.

The feed for the FCCPT stabilization column has been reconstituted on the base of data for the distillation characteristics, the density of the liquid products, and the individual component composition of the gaseous products. Three cases were simulated: Simulation Case 1 using ASTM D 86 for wild naphtha, and diesel distillation data characterization, and simulated distillation ASTM D 2887 extended for the stabilized HTVGO characterization; Simulation Case 2 with the H-Oil VGO blend, and the stabilized HTVGO (the bottom product from the FCCPT stabilization column) distillation characteristics measured by the ASTM D 1160; and Simulation Case 3 with the H-Oil VGO blend, and the stabilized HTVGO distillation characteristics measured by the ASTM D 7169 method. Characteristics of the hydrocarbon gas product, the wild naphtha, the diesel fraction, and the stable hydrotreated VGO for the three studied simulation cases are presented in Tables S1–S4. It is evident from the data in the Tables S1–S3 that the distillation characteristics of the wild naphtha, the diesel fraction, and the stabilized hydriotreated VGO (HTVGO) are almost the same, that is an indicator of the same efficiency of fractionation for all studied cases.

It is interesting to note that the sulphur content in the diesel faction and the HTVGO is higher in Case 2 due to the higher sulphur content in the H-Oil VGO blend. This is also an indicator that the diesel fraction in Case 2 contains material from the H-Oil VGO blend.

The outlet stream summation for Simulation Case 1 is shown in Figure S1. The FCCPT stabilization section simulation for heat and material balance for Simulation Case 1 is represented in Figure S2, where the calculated temperature, pressure, and mass flowrate for each material stream is displayed along with the readings from the instruments. It can be seen from this data that the simulation model is a close representation of the actual operation of the FCCPT stabilization section.

In Figure S3 the summation of the FCCPT stabilization column outlet streams to form the inlet stream composition is done, while the characterization of the H-Oil VGO blend and the stabilized HTVGO streams is based on the ASTM D1160 analyses. The following assumption was made for Simulation Case 2: since the flowrate and characteristics of the feedstock to FCCPT stabilization section are the same for Simulation Case 1 and Simulation Case 2, the same yield of the diesel fraction is assumed for Simulation Case 1 and Simulation Case 2 with the same distillation curve.

Figure S4 represents the heat and material balance performed for Simulation Case 2, using the above-mentioned ASTM D 1160 analyses to sum-up the inlet stream to the FCCPT stabilization section. It is visible that the flow rate of the diesel outlet stream from the side-stripper K-102 deviates considerably from the amount measured with the field flowmeter.

Figure S5 shows the summation of the outlet streams of the FCCPT stabilization section, but now the characterization of the H-Oil VGO blend and the HTVGO are defined using the results from the ASTM D 7169 analyses. It shall be noted that the ASTM D 7169 distillation curves were entered as TBP curves directly. The same applies for the flowrate and distillation properties of the diesel fraction in this version of Simulation Case 3. Figure S6 displays the heat and material balance for Simulation Case 3, where the inlet stream uses the VGO streams characterization by HTSD ASTM D 7169 method. It can be seen that the material balance here closely matches the actual plant operation. The diesel draw-off temperature is slightly higher than the actual instrument readings and could be attributed to deviations of the predicted essential properties of the heavier (360 °C+) pseudo-components, such as molecular weight and critical properties.

4. Discussion

The proper characterization of petroleum oils is important not only for distillation column design but also for correct evaluation of the existent separation equipment performance [26,27,28]. Kaes [25] states that an ASTM D 1160 apparatus can be used to extend the crude TBP data. He pointed out that the ASTM D 1160 distillation better approximates a TBP distillation because it is carried out under vacuum; can distill heavier materials than a TBP; much simpler to run than TBP; and it is the best test for refinery vacuum distillation unit calculations [25]. Kaes considers the gas chromatographic simulated distillation method to be an unreliable method for conversion to TBP curve, and due to the fact that it derives the boiling data by using standard mixtures of n-paraffins, it is deemed to report inaccurate results for streams with significant aromatic content [25].

Golden et al. [26,27,28], however, argue that the use of the ASTM D 1160 method is worse than employing inaccurate TBP data in the design process of a vacuum distillation column. They conclude that few designers know the proper feed characterization techniques, including the use of a modern ASTM D 7169 chromatographic method. Golden et al. summarize that more than 50 vacuum distillation units have been field tested to confirm the validity of using ASTM D 7169 as a TBP curve to model crude and vacuum distillation unit operation [35].

While Kaes [25] seems to prefer using ASTM D 1160 for heavy oil characterization, Golden et al. [26,27,28] categorically deny ASTM D 1160, and strongly recommend the use of ASTM D 7169 for characterization of heavy oils.

The data shown in Figure 5 appear similar to the comparison of distillation curves of the ASTM D 86 and the simulated distillation (ASTM D 2887) reported by Riazi [30]. The data in Figure 5 also look similar to the ASTM D 86 and TBP curves reported by Riazi [30] Both the TBP and simulation distillation curves lie below the ASTM D 86 and the ASTM D 1160 curves for evaporation lower than 50% and lie above these curves for evaporation higher than 50%. Riazi [30] argues that the ASTM D 86 distillation data do not represent actual boiling point of components in a petroleum fraction, while the TBP, due to the high degree of fractionation, gives accurate component distribution for mixtures. Riazi also concludes that ASTM D 1160 distillation data is measured more accurately than ASTM D 86 since it is conducted at low pressure [34]. For this reason, ASTM D 1160 curves are closer to TBP curves at the same pressure base. Nevertheless, the similar appearance of the difference in the distillation curves between TBP and ASTM D 86 and that observed in the data of Figure 5 for ASTM D 7169 and ASTM D 1160 may suggest that ASTM D 7169 could be closer to TBP than the ASTM D 1160. Moreover, in our earlier research [24], we proved that the simulated distillation method ASTM D 2887 is equivalent to TBP, even for high aromatic oils like the fluid catalytic cracking light and heavy gas oils. All these findings suggest that the gas chromatographic HTSD (ASTM D 7169) should be closer to the TBP than the physical vacuum distillation method ASTM D 1160, and therefore, it should better characterize the heavy petroleum fractions.

The data in Table 2 show that the LNB FCC feed hydrotreating section operated at the same conditions in the hydrotreating reaction section (the same reaction temperature = 357 °C; the same feed rate = 160 t/h; and the same reaction pressure = 41 bar (g)) for both studied cases. Considering the same quality of the feed for the LNB FCC feed hydrotreating section (see the data in Table S5), one can expect the same vacuum gas oil conversion to take place in the hydrotreating reaction section, and thus, the same amount of generated diesel fraction in this section for both studied cases. Therefore, a different diesel fraction quantity registration in both cases must be ascribed to a separation of an additional amount of diesel fraction from the H-Oil VGO blend processed in the stabilization column during Case 2. The higher fuel gas consumption in the heater, heating the stabilization column feed during Case 2 suggests a higher extent of evaporation occurring in the stabilization column feed during Case 2, which could be due to a higher amount of low boiling material in the column feed in Case 2. If one calculates the total fluid flow rate through the six passes of the heater from Figure 6, it can be seen that for Case 1, it is 562 m³/h, versus 552 m³/h for Case 2. Irrespective of the lower total fluid flow rate in the heater in Case 2, the higher fuel gas consumption confirms the suggestion that the stabilization column feed should contain a higher amount of lighter material. The higher total fluid flow rate in Case 1, irrespective of the lower throughput (160 t/h feed rate for the FCCPT in Case 1) than in Case 2 (160 t/h feed rate for the FCCPT reaction section plus 39 t/h H-Oil VGO blend in the stabilization section), can be explained by the higher percent of recycling of stabilized HTVGO through the heater. This is evident from the higher percent of opening the valve providing recycling during the heater for Case 1.

The data in Table 3 and Table 4 indicate much better agreement between the separated diesel fraction from H-Oil VGO blend (8.7 wt.%) in the FCCPT stabilization column and that measured in the H-Oil VGO blend by ASTM D 7169 (9.6 wt.%) than the measured by ASTM D 1160 diesel fraction content in the H-Oil VGO blend (1.7 vol.%). Having in mind that the density of the diesel fraction (see the data in Table S2) is lower than that of the H-Oil VGO blend, the amount of diesel fraction in H-Oil VGO blend becomes even lower; 1.6 wt.% based on the ASTM D 1160 distillation data. These results support the opinion of Golden et al. [26,27,28] that the gas chromatographic method ASTM D 7169 better characterizes heavy oils than the ASTM D 1160 method. The separation efficiency of the FCCPT stabilization column seems to be higher than that of the ASTM D 1160 apparatus, and it is better matched to the ASTM D 7169 method.

The data in Table 5 exhibit that the removal of the diesel fraction from the H-Oil VGO has a positive effect on the FCC unit performance leading to increased conversion by 0.53% that along with the increased diesel production in the FCCPT section resulted in improved refinery performance expressed by increased margin with 7.2 MM USD/year.

The results shown in Figure S2 (Simulation Case 1) indicate a very good match between the operating data and the simulated data of the stabilization column. This suggests that the characterization of the streams of the stabilization column using ASTM D 86 distillation data for wild naphtha, and FCCP diesel, and the simulated distillation ASTM D 2887 data for the stabilized hydrotreated vacuum gas oil and the applied approach the FCCPT stabilization column operation simulation are correct.

The results shown in Figure S4 (Simulation Case 2) show that agreement between the operating data and the simulated data of the stabilization column is not very good, especially for the draw-off of diesel fraction. The draw-off of diesel fraction is simulated to be 17,900 kg/h, while the actual draw-off of the diesel fraction was 20,150 kg/h. This is an indicator for underestimation of diesel fraction content in the feed for the stabilization column, which should be attributed to the lower content of diesel fraction in the heavy oils measured by the ASTM D 1160 method for the H-Oil VGO blend, and the stabilized hydrotreated vacuum gas oil. All attempts to increase the draw-off of diesel fraction like modifying operating conditions of the simulated stabilization column failed to give any positive results. This clearly implies that the column feed characterization is not sufficiently well defined when ASTM D 1160 is used.

The results shown in Figure S6 (Simulation Case 3) show that agreement between the operating data and the simulated data of the stabilization column is very good. The draw-off of diesel fraction is simulated to be 20,150 kg/h, which exactly equals the measured diesel fraction draw off. However, this excellent agreement between measured and simulated diesel draw-off has been achieved by a slightly higher than the actual instrument readings draw-off temperature. This can be attributed to deviations of the predicted essential properties of the heavier (360 °C+) pseudo-components, such as molecular weight and critical properties, as was discussed in our earlier study [39]. In opposition to Simulation Case 2, where the column feed characterization was based on ASTM D 1160 distillation data, Simulation Case 3 allowed for a fine-tuning of the column performance simulation, indicating that the ASTM D 7169 is superior to the ASTM D 1160 method. Another important finding in this simulation study is that the ASTM D 7169 entered as a TBP distillation in the software process simulator led to the obtaining of satisfactory simulation results. This implies that the ASTM D 7169 applied for heavy oils (boiling above 360 °C) characterization could be considered equivalent to TBP, as we found out in our earlier research that the simulation distillation method ASTM D 2887 was equivalent to TBP for petroleum oils boiling below 360 °C [24].

5. Conclusions

The distillation characterization of heavy oils is an important petroleum characterization method because it provides information about the quality of separation taking place in the oil refining distillation columns. It is also important for the design of new separation equipment. Laboratory physical vacuum distillation tests (ASTM D 1160) and gas chromatographic simulated distillation analyses are typically used to evaluate the efficiency of separation occurring in the oil refining distillation columns which separate mixtures containing heavy oils. These laboratory methods are reported in the literature to be used for the design of new separation equipment. The results of this study support the opinion expressed by some researchers that the HTSD (ASTM D 7169) is superior for the characterization of heavy oils than the physical vacuum distillation test ASTM D 1160. The commercial test performed at a commercial FCC pretreater stabilization column showed that the yield of diesel fraction separated from a H-Oil VGO blend better matched to the content of this fraction in the H-Oil VGO blend measured by the ASTM D 7169 method than that measured by the ASTM D 1160 method. The better distillation characterization of the heavy oils by the ASTM D 7169 method was also confirmed by the use of software process simulator, demonstrating much better agreement between simulation results with the field observations than those obtained during the use of the ASTM D 1160 method. A conclusion could be made that the ASTM D 7169 method could be considered equivalent to TBP for heavy oils boiling above 360 °C.

Irrespective of the higher costs for single measurement by the ASTM D 7169 method, this method can deliver a substantial improvement in oil refining margin, which outweighs the costs many times over. Thus, we would consider using the ASTM D 7169 method instead of the ASTM D 1160 in our oil refining practice from now on.