Experimental Studies on Diesel Deterioration: Accelerated Oxidation in a Reaction Vessel and Thermogravimetric Analysis

Abstract

Accelerated oxidation experiments on Chinese 0# diesel fuel were performed with a self-designed aging reactor system. Five experimental conditions covering pressures ranging from atmospheric pressure to 0.8 MPa, temperatures ranging from room temperature (25 °C) to 80 °C, and their synergistic effects were adopted to simulate the long-term oxidation of diesel fuel. The extent of deterioration was evaluated based on the measurement of three key indicators, i.e., oxidation stability, wear scar diameter, and viscosity. Thermogravimetric analysis (TGA) tests were performed, and the measured thermogravimetric (TG) curves and derivative thermogravimetric (DTG) curves were used to evaluate the effects of reactor material, heating rate, bath gas, and reactive gas on the deterioration and vaporization processes of diesel fuel. Based on a comparison of the deterioration indicators of diesel fuel collected from the accelerated oxidation experiments and oil depots serving actual operating emergency diesel generators (EDGs), a rapid assessment method of real-time diesel deterioration was explored. Based on the experimental observations, the affecting mechanisms of the increases in temperature and oxygen partial pressure were discussed. Two test methods of accelerated oxidation, with the respective conditions of 0.8 MPa/80 °C and atmospheric pressure/80 °C, were proposed, which could effectively compress the time needed for long-term storage simulations (e.g., 200 h lab aging equals three years of actual operation). The optional temperature and pressure windows for acceleration oxidation were confirmed (40–80 °C/0.3–0.8 MPa). These results are valuable for the further understanding of the processes of deterioration and vaporization of diesel fuel.

1. Introduction

Diesel fuel belongs to the light distillate category of petroleum. It consists of highly complex hydrocarbon mixtures, which can be roughly categorized into paraffin, cycloalkanes, and aromatics [1]. The interdependent components in diesel fuel critically determine its end-use quality, including its combustion and spray characteristics. Some critical parameters of diesel fuel, e.g., viscosity, acidity, lubrication properties, etc., tend to change dramatically after prolonged preservation, which can be attributed to the influence of temperature, oxygen exposure, and external pollutants, such as water, dust, and metal particles [2]. Diesel fuel with a relatively higher degree of deterioration can cause severe engine performance degradation, failures of key parts, or even unplanned shutdown incidents.

Emergency diesel generators (EDGs) serve as critical backup power sources in nuclear facilities, hospitals, and other important infrastructural systems, where fuel storage stability directly impacts operational safety and failure rate. The fuels used in EDGs frequently undergo prolonged storage of up to 10 years, during which many reactions associated with deterioration, including decomposition, oxidation, condensation, etc., can occur. These processes alter key physicochemical properties of diesel fuels, including viscosity (affecting injection performance), oxidation stability (affecting the propagation tendency of radicals), and lubricity, and further affect the reliability of the engine, starting system, and fuel system of an EDG, leading to failures [3] (about 50% of failure events were related to the failures of engines, starting systems, and fuel systems of EDGs). In addition, the Institute for Energy and Transport of the European Commission [4] defines diesel stored for approximately 3 to 6 months or longer as “fuel stored for an extended period of time” and recommends various methods to avoid deterioration.

Unfortunately, experimental and theoretical studies focusing on the deterioration processes of diesel and other fuels similar to it are rarely published. Du Plessis et al. [5] carried out storage tests on surrogate fuels of biodiesel, in which the effects of air, temperature, light, additive, and contact with mild steel were considered. Viscosity was identified as the most evident indicator to evaluate oxidation processes. Bondioli et al. [6] performed an experimental study on the deterioration of biodiesel under controlled storage conditions (at 20 °C and 40 °C, in glass and iron containers). As follow-up studies, they evaluated the storage stability and oxidative stability of fuels based on the ASTM D4625 [7] and ASTM D2275 [8] methods. The production of acids, peroxides, aldehydes, and polymers, and changes in viscosity were identified as deterioration indicators based on experimental observations, while insoluble formations were considered a relatively improper way to evaluate the stability of fuels [9]. The stability of fuels under commercial storage conditions was studied by Bondioli’s group [10], and the duration of experiments lasted for more than one year. It was observed that the viscosity and peroxide value showed significant changes during storage, and occasional shaking could effectively promote the deterioration processes. Middlebach [11] and Serrano et al. [12] mainly studied the influences of antioxidants on the oxidation stability of fuels and reported that the amounts of double bonds of fuel molecules and the fuel reactivity could significantly affect the effectiveness of the additives. Christensen et al. [13] examined the properties necessary for storing biodiesel and biodiesel/diesel blends, and proposed a sample preparation method to simulate the processes of 1-year and 3-year quiescent storage based on ASTM D4625 conditions [7]. Bezergianni et al. [14] performed a 1-year investigation on the oxidation stability of white diesel, a new renewable fuel. A storage condition in the absence of sunlight, at room temperature and under normal atmosphere, was chosen for the deterioration tests, while viscosity, total acid number, induction period, carbonaceous deposits, density, etc. were selected as key deterioration parameters. The experimental results suggested that white diesel maintained good oxidation stability after an extended storage time. Recently, Yu et al. [15] established a gray mathematical model and a GM (1,1) prediction model, which could successfully identify the characteristics of diesel fuel deterioration indications and accurately predict the important indicators of diesel, including the flash point, 50% recovery temperature, and viscosity.

Despite these investigations, fundamental challenges remain unresolved. The current research faces three fundamental limitations:

(1)

Real-time deterioration studies require year-scale monitoring, hindering rapid assessment. However, the studies on the storage and oxidative stability of diesel fuel are still lacking. The optional range for the accelerated oxidation experiment of diesel fuel, including temperature, oxygen exposure, and agitation-related parameters, is still undetermined. In particular, the temperature conditions lower than D2275 (heating bath at 95 °C) for oxidative stability tests should be further explored;

(2)

Existing accelerated tests fail to simulate synergistic property evolution;

(3)

The experimental data of the processes of diesel deterioration, which are affected by the reactor material, heating rate, bath gas, and reactive gas, is still lacking. The vaporization effect during the deterioration was rarely reported.

Notably, water and particulate contaminants, which are primarily introduced through storage tank leakage or refueling, demonstrate extrinsic variability and can be effectively avoided by methods such as optimizing sealing. Hence, these factors were excluded in this study.

Therefore, the accelerated oxidation of diesel was experimentally investigated in a reaction vessel under a wide range of conditions of temperature and oxygen exposure to extend the option range of deterioration tests. Meanwhile, the vaporization processes during the deterioration were studied at the quantitative level and the corresponding effects were determined.

2. Experimental Methods

2.1. Apparatus Design

The accelerated oxidation experiments were performed in a home-made diesel deterioration simulation system. To decrease experimental cycle time and reduce the monitoring intervals, a small reaction vessel with a volume of 5 L was used for the experiments. The diesel deterioration simulation system, the actual image of which is shown in Figure 1, consisted of five core components: (a) an aging vessel mainly containing a 5 L stainless steel tank with a pressure resisting capability up to 1.5 MPa; (b) a heating mantle wrapping the stainless steel tank., which could offer adjustable temperature in the range of room temperature to 150 °C (flexibly controlled by a panel, shown in Figure 1e); (c) a gas flow control system equipped with pressure-bearing sealing elements with good air tightness, used to guarantee the experimental conditions of the flow rate and pressure of oxygen; (d) agitation equipment with an agitation gas source, i.e., a 60 L/0.8 MPa air compressor; and (e) an aging gas source containing a 40 L cylinder of oxygen (>99% purity), an exhaust valve (Figure 1a), pressure-limiting valve (Figure 1b), pressure gauge (Figure 1d), etc. Similar experimental process and instrumentation can be found in Ref. [16].

2.2. Accelerated Oxidation Experiments

In accelerated oxidation experiments, 0# diesel fuel (complying with GB 19147-2016 [17], China’s National Standard, produced by the Sinopec refinery in Tianjin, China) was charged into the aging vessel beforehand. Core components of the diesel fuels used in this work are available in Ref. [18] and key properties, including solid point, density, flash point, etc., can be found in Ref. [19]. The filling volume after fuel charge did not exceed two-thirds of the vessel capacity. Following an air tightness test, the heating mantle was energized, heating the diesel fuel inside the aging vessel gradually to set temperatures. The aging gas, i.e., oxygen, was metered into the aging vessel via a gas flow control system, with steady-state indications on the pressure gauge recorded. Pressure during the accelerated oxidation experiments was controlled by a pressure-limiting valve. The pneumatic agitation was activated when necessary and the pressure was provided by the air compressor.

The detailed experimental conditions were chosen according to the standards ASTM D5304 [20] (Standard Test Method for Assessing Distillate Fuel Storage Stability by Oxygen Overpressure), SH/T 0175 [21] (Standard Test Method for Oxidation Stability of Distillate Fuel Oil (Accelerated Method)), and SH/T 0690 [22] (Standard Test Method for Middle Distillate Fuel Storage Stability at 43 °C). For accelerated oxidation experiments carried out under different temperatures, minimizing the temperature deviations between experimental conditions and the actual storage environment can effectively enhance the accuracy of the diesel deterioration simulation. Considering these above factors, two experimental temperatures, i.e., 40 °C (close to the highest temperature of actual long-term storage of diesel fuel) and 80 °C (close to the temperature of accelerated oxidation experiments of diesel fuel in the standards), were chosen in this work.

To investigate the effects of temperature on diesel deterioration, two representative controlled temperature setpoints, i.e., 40 and 80 °C, were established, spanning the temperatures of actual storage to accelerated oxidation of diesel fuel. Three states were included in the temporal analysis, i.e., (a) baseline unaged diesel fuel, (b) 35 h aged, and (c) 70 h aged. Throughout the deterioration process, continuous pneumatic agitation was maintained in the reaction vessel to ensure uniform reaction conditions. The detailed conditions of the deterioration tests performed at discrete temperatures are given in

Table 1. Detailed conditions of the deterioration tests performed at discrete temperatures.

No.	Gauge Pressure (MPa)	Total Pressure (MPa)	Temperature (°C)	Time (h)
TEST 1	0	0.101325	40	0
TEST 2	0	0.101325	40	35
TEST 3	0	0.101325	40	70
TEST 4	0	0.101325	80	0
TEST 5	0	0.101325	80	35
TEST 6	0	0.101325	80	70

.

To investigate the effects of oxygen partial pressure on diesel deterioration, two representative controlled pressure setpoints (internal pressure of the reaction vessel, rather than oxygen partial pressure), i.e., 0.3 MPa (approximately 0.2 MPa higher than atmospheric pressure, simulating storage tank headspace) and 0.6 MPa (close to the pressure of accelerated oxidation experiments of diesel fuel in the standards), were established, spanning the pressure conditions of actual storage to acceleration oxidation of diesel fuel. To avoid unexpected temperature-dependent reactions, the temperatures for these deterioration tests were set at 25 °C. Similar to the experiments focusing on the effects of temperature, three states, including unaged, 35 h aged, and 70 h aged samples, were included in the temporal analysis. The detailed conditions of the deterioration tests performed under discrete oxygen partial pressure conditions are given in

Table 2. Detailed conditions of the deterioration tests performed under discrete oxygen pressure conditions.

No.	Gauge Pressure (MPa)	Total Pressure (MPa)	Temperature (°C)	Time (h)
TEST 7	0.199	0.3	25	0
TEST 8	0.199	0.3	25	35
TEST 9	0.199	0.3	25	70
TEST 10	0.499	0.6	25	0
TEST 11	0.499	0.6	25	35
TEST 12	0.499	0.6	25	70

.

To tentatively investigate the joint effects of temperature and pressure on the deterioration of diesel fuel, experiments were designed (

Table 3. Detailed experimental conditions of the investigations of the joint influences of temperature and pressure.

No.	Gauge Pressure (MPa)	Total Pressure (MPa)	Temperature (°C)	Time (h)
TEST 13	0.699	0.8	80	0
TEST 14	0.699	0.8	80	35
TEST 15	0.699	0.8	80	70
TEST 16	0.699	0.8	80	200

). To eliminate gas–liquid interfacial disturbance causing pressure fluctuations during pressurizing accelerated oxidation processes, agitation was not adopted in these static-configuration experiments, and only the effects of oxygen partial pressure and temperature were considered. Three states, identical to those in discrete temperature/pressure tests, were included for temporal analysis.

2.3. Oxidation Stability Tests

The oxidation stability tests performed in this work, which quantitatively analyzed insoluble oxidation products from accelerated oxidation experiments and evaluated the formation tendency of gum and deposit, strictly followed the requirements of the standard SH/T 0175 [21]. The procedure of oxidation is briefly introduced below.

The liquid samples (approximately 400 mL) that came from the deterioration tests were filtered under vacuum (about 0.08 MPa). After pre-filtration, 350 mL ± 5 mL samples were oxidized in a clean tube reactor by 50 mL/min ± 5 mL/min oxygen purging at 95 °C ± 0.2 °C for 16 h. The tube reactor was heated in an oil bath to ensure uniform heating conditions. Then, the reaction tube was cooled at room temperature for, at most, 4 h, and the liquid samples were filtered under 0.08 MPa vacuum. To accurately weigh the filterable insoluble, the solid deposit on the filter membrane and the reaction tube were both rinsed with iso-octane three times, and the filtrate was further filtered under vacuum. The deviations between the filter membrane used for the above-mentioned operations and an untreated baseline membrane were recorded as the mass of filterable insolubles.

The adherent deposits on the inner wall of the reaction tube were dissolved with 75 mL ± 5 mL tri-solvent (the volume fractions of acetone, methanol, and toluene equaled 1:1:1) by rinsing three times. The washing fluids were transferred to a 200 mL beaker and vaporized at 135 °C. As a blank control, untreated tri-solvent was vaporized under the same conditions. The mass deviations between the undissolved substances from control and experimental groups were recorded as the masses of adherent insolubles.

The sum of the masses of filterable and adherent insolubles was recorded as total insolubles, the key evaluation index of the oxidation stability of diesel fuel.

2.4. Wear Scar Diameter and Viscosity

The diesel lubricity tests performed in this work, which evaluate the extent of diesel deterioration for EDG oil depots, strictly followed the requirements of the high-frequency reciprocating rig (HFRR) method in the standard NB/SH/T 0765-2021 [23].

During the diesel lubricity tests, 2 mL ± 0.2 mL liquid sample was placed in the test reservoir of the HFRR. A vibrator arm holding a non-rotating steel ball (6 mm diameter, high-carbon chromium bearing steel) was lowered until contacting a test disk (high-carbon chromium bearing steel) completely submerged in the liquid sample. The vibrator was loaded with a 200 g weight to provide normal force. When the temperature of the sample in the oil reservoir was stabilized at 60 °C ± 2 °C, the steel ball rubbed against the disk at 50 Hz ± 1 Hz frequency and 1 mm ± 0.02 mm stroke for 75 min ± 0.1 min (set values). The images of wear scars were captured with a microscope digital camera, and measured along X- and Y-directions with an accuracy of 10 μm.

The diesel viscosity tests performed in this work strictly followed the National Standard GB/T 265-1988 [24]. The experimental method has been widely used and, thus, no more detailed description is given.

2.5. Thermogravimetric Analysis (TGA)

A METTLER TOLEDO 2 thermal analysis system was employed in TGA tests. The specifications of the thermal analyzer are summarized below: (a) equipped with a METTLER TOLEDO ultra-micro balance (1 μg resolution); (b) analyzing samples from room temperature up to 1200 °C; (c) measuring and recording up to 50 million data points continuously; and (d) containing built-in gas flow control system containing three gas paths, i.e., protective gas, reactive gas, and purge gas. The gas flow control system enables testing under various atmospheres, including inert gases (e.g., argon) and reactive gases (e.g., air and oxygen).

Approximately 40 μL diesel samples were transferred into crucibles using a precision pipette (with a range of 20~50 μL): 80 μL sapphire crucibles were adopted for most TGA tests due to good chemical inertness, while Pt-Rh crucibles with the same volumes were exceptionally used for reactor material effect studies. Two taring operations, before and after placing the empty crucibles on the balance, were performed during sample processing to ensure the recorded mass represented solely sample mass, excluding the crucible. Crucibles were carefully positioned on the sample holder of the ultra-micro balance using ceramic-tipped tweezers. After closing the furnace lid, temperature-programmed heating initiated when temperature stabilized at target initial temperature. Thermogravimetric (TG) curves, which represent the processes of continuous mass loss with the increase in temperature, were automatically recorded by the thermal analysis system. Similar to the previous studies by our group [25], derivative thermogravimetric (DTG) curves in this work are defined as the derivative of absolute sample mass to time, representing mass loss rates.

Test repeatability was evaluated previously [25,26], with stability verified as excellent, which is in good agreement with the verification experiment performed in this work (Figure 2). Blank curves were measured using empty crucibles in the temperature range from room temperature to 500 °C, revealing a maximum error of only 30 μg, which is insufficient to affect TG/DTG distributions.

To investigate the effects of reactor material, heating rate, bath gas, gas exposure, and reactive gas on the vaporization of diesel fuel, experiments following different temperature programs and under various conditions were performed, and the detailed experimental conditions are given in

Table 4. Detailed experimental conditions for TGA tests.

No.	Focus	Reactor	Method	Heating Rate	Bath Gas
Test 1	Reactor material	Pt-Rh crucible	Open	2 °C/min	Ar 20 mL/min
Test 2		Pt-Rh crucible	Sealed	2 °C/min	Ar 20 mL/min
Test 3		Sapphire crucible	Open	2 °C/min	Ar 20 mL/min
Test 4		Sapphire crucible	Sealed	2 °C/min	Ar 20 mL/min
Test 5	Heating rate	Sapphire crucible	Open	5 °C/min	Ar 20 mL/min
Test 6		Sapphire crucible	Sealed	5 °C/min	Ar 20 mL/min
Test 7		Sapphire crucible	Open	10 °C/min	Ar 20 mL/min
Test 8		Sapphire crucible	Sealed	10 °C/min	Ar 20 mL/min
Test 9	Bath gas	Sapphire crucible	Open	5 °C/min	Ar 0 mL/min
Test 10		Sapphire crucible	Open	5 °C/min	Ar 40 mL/min
Test 11		Sapphire crucible	Sealed	5 °C/min	Ar 40 mL/min
Test 12	Reactive gas	Sapphire crucible	Open	5 °C/min	Air 20 mL/min
Test 13		Sapphire crucible	Sealed	5 °C/min	Air 20 mL/min

.

2.6. Actual Diesel Engine Operating and Sample Collection

Diesel samples were collected every six months from EDG oil depots at Hongyanhe Nuclear Power Station (Dalian, China). The storage of the diesel fuel ranged from 0 to 1247 days. For samples collected from 12 EDG oil depots (1P, 1Q, 2P, 2Q, 3P, 3Q, 4P, 4Q, 5P, 5Q, 6P, and 6Q), viscosity and wear scar diameters were measured with the methods mentioned in Section 2.3 and Section 2.4. Acidity, density, and cetane number were, respectively, measured according to standards GB/T 258-2016 [27], GB/T 1884-2000/1885-1998 [28,29,30], and GB/T 386-2021 [31].

3. Results and Discussion

3.1. Experimental Investigations in the Reaction Vessel

The accelerated oxidation results are illustrated in Figure 3, Figure 4 and Figure 5, and the results of the tests of oxidation stability, wear scar diameter, and viscosity are given in Figure 3, Figure 4 and Figure 5, respectively.

Figure 3 shows increased oxygen partial pressure and temperature can both promote diesel oxidation, as oxidation stability increases continuously with temperature for all the experimental conditions. When the time reached 70 h, the oxidation stability increased by 24.6%, 41.5%, 44.6%, 129.2%, and 403.1% for the samples deteriorated under conditions I–V. Conditions I–V represent conditions of 0.3 MPa/25 °C, 0.6 MPa/25 °C., atmospheric pressure/40 °C, atmospheric pressure/80 °C, and 0.8 MPa/80 °C, respectively. The experimental results indicate that temperature effects exceed oxygen partial pressure effects. Under the joint effects of high-temperature and high oxygen partial pressure, diesel oxidation can be significantly accelerated, resulting in a deterioration extent that is up to 16 times that of conditions I–IV. However, the promotion effects with the increases in temperature and pressure will eventually reach a plateau according to the experimental observations, as a clear deceleration in the oxidation stability increase rate is observed for condition V in the time range of 70–200 h.

Figure 4 shows the lubricity improvement during oxidation for conditions I–V, with observable wear scar diameter decreases. The wear scar diameters, respectively, decreased to 99.5%, 98.7%, 98.0%, 96.2%, and 94.4% for conditions I–V when the time reached 70 h. Similar to the effects on the oxidation stability, the increases in temperature and pressure both have promotion effects on the lubricity of diesel fuel, and temperature effects exceed oxygen partial pressure effects. The promotion effects on the lubricity, to some extent, can be attributed to the formation of polycyclic aromatic hydrocarbons (PAHs), and nitrogen- and oxygen-containing heterocyclic polar compounds during the oxidation processes [32,33]. However, the promotion effects are limited under the conditions of relatively lower temperature and oxygen partial pressure when compared with those observed for condition V. For condition V, the promotion effects significantly strengthened with time, reaching a deterioration extent approximately six times that of condition IV, indicating that a suitable design of the joint effects of temperature, oxygen partial pressure, and deterioration time can effectively accelerate the oxidation of diesel fuel and a joint temperature/pressure/time optimization extends the optional experimental range for accelerated oxidation.

Figure 5 shows the viscosity of diesel fuel with time for conditions I–V. Compared with the results of the oxidation stability tests, the changes in viscosity are limited: 0.3%, 0.8%, 1.5%, 2.3%, and 4.1% increases in viscosity relative to the values of baseline unaged diesel fuel were, respectively, recorded after 70 h of accelerated oxidation experiments. The promotion effects of the temperature increases (1.5% → 2.3%) are similar to those of the increases in oxygen partial pressure (0.3% → 0.8%), while the experiments performed at relatively higher temperatures (conditions III and IV) show more significant deterioration effects compared with those of pressuring accelerated oxidation. The promotion effects may originate from PAHs and other high-viscosity larger molecules formed during the accelerated oxidation processes. In the range of 0–200 h, the continuous environment of high temperature and high oxygen partial pressure could effectively promote the increase in viscosity, leading to an 8.7% increase when the time reached 200 h. An abnormal decrease is observed at 35 h for condition III, which can be caused by measurement error or the physically dominated process involving increased molecular kinetic energies and weakened intermolecular forces. With the increase in temperature, the chemically dominated process involving the formation of high-viscosity compounds becomes dominant and promotes the increase in viscosity, which is in good agreement with the results for conditions III and IV. The experimental results indicate that temperatures under 40 °C should be avoided for accelerated oxidation experiments due to the limited deterioration extents and viscosity fluctuations.

In summary, temperature and oxygen partial pressure increases can both promote diesel oxidation, resulting in changes in three key deterioration indicators, i.e., oxidation stability, wear scar diameters, and viscosity. The mechanisms of the above-mentioned promotion effects are speculated below: Increased oxygen partial pressure leads to the enhancement in free oxygen dispersed in diesel fuel. The increase in the oxidation contact area can, to some extent, promote the oxidation processes. On the other hand, increased temperature accelerates molecular activity and hydrocarbon (or other molecules) oxidation rates. Meanwhile, for oxidation stability and lubricity, the promotion effects of temperature are relatively larger than those of oxygen partial pressure because thermal activation dominates radical generation. The joint effects of high temperature and high oxygen partial pressure lead to much more effective deterioration processes.

Researchers are supposed to implement rigorous safety measures prior to replicating the high-pressure/high-temperature protocols mentioned in this work due to the following risks:

(1)

The explosion potential at a total pressure higher than 0.3 MPa—the pressurization should not be performed before reaching the target temperatures, and the pressure ramp rate should be limited according to engineering practice;

(2)

The autoignition risk of diesel vapor (higher than their flash points);

(3)

The fire hazard in the absence of an active quench system and continuous monitoring;

(4)

Oxygen and other forms of corrosion at high temperatures, including oxygen permeation and PAH adsorption.

3.2. TGA

TG and DTG curves are illustrated in Figure 6, Figure 7, Figure 8 and Figure 9, and the results of the tests focusing on the effects of reactor material, heating rate, bath gas, and reactive gas on the vaporization of diesel fuel are given in Figure 6, Figure 7, Figure 8 and Figure 9, respectively. T10 (temperature at 10% evaporation during a TGA test, similar for T50 and T90), T50, and T90 are summarized in

Table 5. The T10, T50, and T90 for TGA tests.

No.	Focus	Reactor	Method	T10 (°C)	T50 (°C)	T90 (°C)
Test 1	Reactor material	Pt-Rh crucible	Open	98.73	158.33	208.40
Test 2		Pt-Rh crucible	Sealed	113.27	168.43	231.27
Test 3		Sapphire crucible	Open	98.13	155.20	204.93
Test 4		Sapphire crucible	Sealed	131.30	209.20	264.50
Test 5	Heating rate	Sapphire crucible	Open	118.17	180.17	231.50
Test 6		Sapphire crucible	Sealed	173.33	240.25	291.92
Test 7		Sapphire crucible	Open	134.50	197.17	248.83
Test 8		Sapphire crucible	Sealed	188.17	254.17	306.17
Test 9	Bath gas	Sapphire crucible	Open	122.08	184.75	235.33
Test 10		Sapphire crucible	Open	119.33	182.67	234.75
Test 11		Sapphire crucible	Sealed	122.58	200.25	280.92
Test 12	Reactive gas	Sapphire crucible	Open	116.67	178.67	229.17
Test 13		Sapphire crucible	Sealed	171.50	238.25	293.42

. All samples underwent TGA tests with open pans and sealed pans with pin-holed lids under similar experimental conditions. The shapes of the main body and the pin-holed lid of the sapphire crucible are available in Figure 10.

Figure 6 and Table 5 show that TG and TGA curves measured with open pans exhibit slight discrepancies (the maximum discrepancies between T10, T50, and T90 are below 4 °C), indicating that the crucible thermal conductivity and catalyzing effects of platinum result in a negligible enhancement in the volatility of diesel fuel in a relatively open environment with thorough contact with the purging gas. However, the good thermal conductivity and catalyzing effects of Pt-Rh crucibles significantly promote the volatility in sealed pans, shifting the TG and DTG curves to a lower-temperature region, and reducing T10, T50, and T90 from 131.30, 209.20, and 264.50 °C to 113.27, 168.43, and 231.27 °C. DTG fluctuation regions can be observed at approximately 150 °C for the sealed Pt-Rh crucible and at approximately 125 °C for the sealed sapphire crucible due to the pressure fluctuation in the enclosed environment. For identical crucibles, the vaporization onset and endset temperatures of the diesel fuel in the open pans (1% conversion at 54.67 °C and 99% conversion at 228.73 °C; Test 1) are much lower than those in the sealed pans (1% conversion at 69.37 °C and 99% conversion at 253.07 °C; Test 2) due to the enhanced volatilization driving forces in open pans, which is in good agreement with previous studies [34]. This phenomenon can be observed in Figure 7, Figure 8 and Figure 9, and, thus, it will not be further elaborated hereinafter.

The effects of heating rates on the vaporization of diesel fuel are illustrated in Figure 7. It shows that higher heating rates shift the vaporization temperature range, as well as the onset and endset temperatures of diesel vaporization, towards higher-temperature regions. T10, T50, and T90 of Test 3 are 98.13, 155.20, and 204.93 °C, increasing to 118.17, 180.17, and 231.50 °C in Test 5 and 134.50, 197.17, and 248.83 °C in Test 7. Sealed crucibles show similar trends (Test 4, 6, and 8). The measured DTG curves show similar halfwidths of the DTG peaks, indicating that the heating rates only have slight effects on the vaporization temperature ranges. The peak vaporization rates clearly increase with temperature. It is worth noting that the DTG curves after the peak are relatively flat, reflecting the long vaporization temperature ranges and poor volatility of heavy components in diesel fuel and high-viscosity deterioration products.

Figure 8 shows the TG and DTG curves of diesel fuel under different conditions of flow rates of bath gas. It shows argon flow rates negligibly affect vaporization in open pans (for Tests 5, 9, and 11, T10 equals 118.17, 122.08, and 119.33 °C; T50 equals 180.17, 184.75, and 182.67 °C; and T90 equals 231.50, 235.33, and 234.75 °C); only very slight promotion effects are observed. At a smaller bath gas rate (20 mL/min), the vaporization range measured with TG and DTG curves was almost monolithically shifted towards the higher-temperature region when the sealed pan was used (for Tests 6 and 11, T10 equals 173.33 and 122.58 °C; T50 equals 240.25 and 200.25 °C; and T90 equals 291.92 and 280.92 °C). A higher bath gas rate (40 mL/min) obviously promotes light/heavy fraction separation due to the enhanced vapor diffusion under high-flow conditions, and, thus, a two-peak phenomenon is observed in the DTG curve. For light components of the diesel fuel, the higher bath gas rate effectively reduces the discrepancy between the vaporization processes in the open and sealed pans, leading to similar TG and DTG curves at temperatures under 150 °C. However, the promotion effects on the vaporization of heavy components are still limited even under high-flow conditions.

The effects of reactive gas (air) on the vaporization of diesel fuel are illustrated in Figure 9. The TG (Figure 9a) and DTG curves (Figure 9b) show that limited suppressing effects could be observed when the reactive gases were used, whether in the open pans or in the sealed pans, which is in good agreement with the measured T10, T50, and T90. The suppressing effects can be attributed to two factors: (a) air’s higher thermal conductivity reducing local temperatures and vaporization rates; and (b) oxidation reactions slowing the apparent weight loss.

The actual images of the sapphire crucibles after TGA Tests 9, 10, 13, and 12 are available in Figure 10. It shows the crucible after Test 13 (Figure 10c) has the most severe condition of carbon residues and the crucible after Test 10 (Figure 10b) resembles an unused one, indicating an enclosed environment, low bath gas rate, and reactive gas enhance carbon residue formation, while good ventilation can inhibit this in an atmosphere of reactive gases.

3.3. Actual Diesel Engine Operating

The values of the acidity, wear scar diameter, viscosity, cetane number, and density of the diesel fuel collected from the EDG oil depots are available in Figure 11. The box plots reveal significant deviations in the acidity, wear scar diameter, and cetane number relative to both the median and the mean, indicating the evident deterioration of the diesel fuel during actual operating. Among them, the increase in acidity can be attributed to the acidic compounds produced via diesel oxidation, whose mechanism is similar to that of the oxidation stability increase [35,36,37]. The increase in cetane number can be attributed to the evaporation of the components with high vapor pressures and low cetane numbers, the oxidation of aromatic hydrocarbons, or the changes in the relative content of the n-alkanes [38], which counteract the PAH effects. Figure 11c,d show that the changes in the viscosity and density of diesel fuel were relatively limited during the actual storage, which is in good agreement with the accelerated oxidation results in this work.

To determine the temporal variation patterns of the key deterioration indicators, i.e., the wear scar diameter and viscosity, linear fittings were performed, and the results are given in

Table 6. Slopes and R² of linear fittings between storage data (wear scar diameters and viscosity) and storage days.

Depot No.	Storage Days	Linear Fitting (Wear Scar Diameter)		Linear Fitting (Viscosity)
		k	R2	k	R2
1P	1149	0.02372	0.10	3.27 × 10−5	0.050
1Q	967	0.01861	0.029	1.63 × 10−4	0.13
2P	1181	−0.08585	0.70	−5.15 × 10−5	0.055
2Q	1241	0.02422	0.14	1.90 × 10−5	0.16
3P	1065	−0.01155	0.011	2.89 × 10−4	0.45
3Q	1247	0.00249	0.0011	8.76 × 10−5	0.30
4P	1148	−0.04004	0.12	1.42 × 10−4	0.57
4Q	1135	−0.08995	0.31	1.43 × 10−4	0.81
5P	351	−0.01997	0.022	2.62 × 10−4	0.54
5Q	351	−0.00619	0.0031	2.58 × 10−4	0.92
6P	258	0.03393	0.72	5.43 × 10−5	0.22
6Q	258	−0.15976	0.79	2.31 × 10−4	1.00

. The R² values show very poor linear correlations between the deterioration indicators and operating days for the diesel fuel collected from most of the EDG oil depots (<0.7 in 75% samples for wear scar diameter and viscosity), indicating that the contributing factors in diesel deterioration for actual storage can be very complex and suggesting that factors beyond temporal influence dominate actual deterioration. Figure 12 compares the deterioration effects in accelerated oxidation experiments and actual storage in EDG oil depots, in which the deterioration indicators of the diesel fuel collected from two representative EDG oil depots (with an R² higher than 0.7), i.e., wear scar diameter for EDG oil depot 2P and viscosity for EDG oil depot 4Q, are illustrated.

For the diesel fuel collected from EDG oil depot 2P, the wear scar diameter decreased by 24.3% after 1181 days of actual storage. And the wear scar diameter significantly decreased by 27.7% after 200 h of accelerated oxidation for condition V. The viscosity of the diesel fuel collected from EDG oil depot 4Q increased by 7.3% after 1091 days of actual storage, while the viscosity after 200 h of accelerated oxidation (condition V) exhibits an increase of 8.7%. The comparison between accelerated oxidation experiments and actual storage indicates that the accelerated oxidation method designed in this work, i.e., condition V, can simulate the deterioration of diesel fuel after actual storage for approximately three years. Similarly, condition IV can simulate the deterioration of diesel fuel after the actual storage in EDG oil depots for at least one year based on the viscosity comparison.

4. Conclusions

Based on a self-designed aging reactor system of diesel deterioration, accelerated oxidation experiments were performed under five simulation conditions, including increases in temperature and oxygen partial pressure, and their joint effects. The deterioration extents were evaluated based on the measurements of oxidation stability, wear scar diameter, and viscosity. TGA tests were performed to evaluate the effects of reactor material, heating rate, bath gas, and reactive gas on the deterioration and vaporization processes of diesel fuel. The rapid assessment method of real-time diesel deterioration was explored by comparing the lab/actual diesel properties. The results of this work lead to the following conclusions:

(1)

A temperature increase exhibits significantly stronger promotion effects on the deterioration of diesel fuel than an oxygen partial pressure increase. High temperatures promote diesel deterioration by activating radical-chain reactions, while the promotion effects of elevated oxygen partial pressure are attributed to the enhancement in oxygen mass transfer.

(2)

Non-temporal factors dominate actual diesel deterioration for EDG oil depot storage. Weak linear correlations between the time and viscosity/wear scar diameter were discovered.

(3)

A test method of accelerated oxidation under conditions of static 0.8 MPa and 80 °C was proposed, which could effectively compress long-term storage simulation (200 h lab aging equals three years of actual storage, summarized based on the comparison of the data of the wear scar diameter and viscosity obtained from accelerated oxidation and several representative oil depots). The optional temperature and pressure windows for acceleration oxidation were confirmed (40–80 °C/0.3–0.8 MPa).

(4)

The database of the deterioration and vaporization of diesel fuel was extended. TGA tests revealed critical vaporization controls, including the methods for suppressing carbon residues and lowering vaporization thresholds.