UNIVERSITATEA PETROL-GAZE DIN PLOIETIFACULTATEA DE INGINERIE MECANIC I ELECTRICINGINERIA EXPLOATRII OPTIMALE A UTILAJULUI PETROLIERMETODOLOGIA CERCETRII TEORETICE I EXPERIMENTALE

STUDIU DE CAZ

Bio-Oil Properties and Effects on Containment Materials

ef lucr. dr. ing. Doina PETRESCUMasteranzi:Marian-Valentin MITREAIonu-Florentin SANDU

PLOIETI2015Bio-Oil Properties and Effects on Containment Materials(Proprietile bio-combustibilului i efectele sale asupra materialelor n care se depoziteaz)Ionu-Florentin SANDUMarian-Valentin MITREAUniversitatea Petrol-Gaze din PloietiREZUMATLichidele obinute din biomas ofer un potenial pentru transformare n combustibili lichizi cum ar fi benzina, cherosenul, motorina i pcura. Totui, bio-combustibilii, creai prin oricare dintre numeroasele procese, conin concentraii importante de ap i compui organici n componena crora se afl oxigen. Acizii carboxilici constituie o parte semnificativ din combustibilii care conin oxigen i acesti compui, n special acizii formic i acetic, fac bio-combustibilii destul de acizi. Daca acizii carboxilici nu sunt nlturati, bio-combustibilul poate crea probleme de coroziune pentru materialele folosite n construirea sistemelor utilizate n producia, procesarea, transportul i stocarea acestuia.Au fost dezvoltate tehnici care mbuntesc caracteristicile bio-combustibililor i studii de laborator legate de coroziune, fcute att cu pri organice, ct i acvifere, pentru a evalua coroziunea substanelor obinute din biomas. Pe lng aceasta, mostrele au fost expuse n sisteme de lucru i componentele acestor sisteme, mpreun cu mostrele expuse, au fost examinate pentru a evalua msura degradrii. Rezultatele acestor studii de caracterizare vor fi descrise i se vor face recomandri pentru materiale suficient de rezistente n aceste medii.Cuvinte cheie: Bio-combustibil, coroziune, acizi carboxilici, acid formic, acid acetic, cromatografia gazelor, electroforez capilar.ABSTRACTBiomass derived liquids offer a potential for conversion into liquid fuels such as gasoline, jet fuel, diesel, and fuel oil. However, the bio-oils, whether produced by any of a number of processes, contain significant concentrations of water and oxygen-containing organic compounds. Of the oxygenates, carboxylic acids constitute a significant fraction, and these compounds, primarily formic and acetic acids, make bio-oil quite acidic. Unless the carboxylic acids are removed, bio-oil presents corrosion issues for materials used to construct the systems used for production, processing, transport and storage of the bio-oil.Techniques have been developed to improve characterization of the bio-oils, and laboratory corrosion studies have been conducted with both the organic and aqueous fractions to assess the corrosivity of the biomass-derived materials. In addition, samples have been exposed in operating systems, and components of these systems, as well as the exposed samples, have been examined to assess the extent of degradation. The results of these characterization studies will be described and recommendations for materials sufficiently resistant to these environments will be provided. Key words: Bio-oil, corrosion, carboxylic acids, formic acid, acetic acid, gas chromatography, capillary electrophoresis.INTRODUCTIONFor both economic and environmental reasons, there is a major effort around the globe to make greater use of renewable energy resources. Almost certainly, a variety of renewable resources will be utilized, and biomass is expected to be a significant contributor in many parts of the world including the United States. Many processing approaches are being utilized, including combustion of unprocessed biomass as well as biomass that has been treated to produce a gaseous, liquid or solid fuel. Each processing method has advantages and disadvantages, and many studies are underway to overcome the disadvantages and improve the efficiency of each of these processes. One issue that is common to all the processing methods is the degradation, generally through corrosion, of the vessels and other components that are used to contain the processes.One of the current authors has published several papers addressing the corrosion issues that have been encountered in production of a gaseous product whether by gasification or steam reforming. Several of the current authors have published a number of reports describing studies to address materials in fast pyrolysis or other means to produce a biomass-derived liquid. These reports describe the analytical techniques developed and used to identify the corrosive components in the biomass-derived liquids. Laboratory corrosion studies have also been conducted by the authors to determine the effect of bio-oils on selected alloys typically used to construct chemical processing systems. The unavailability of appreciable quantities of some bio-oils as well as the concern that tests conducted with a limited volume of bio-oil could be influenced by depletion of the corrosive component have led to efforts to expose samples in biomass processing systems as well as to examination of components from operating systems. In an effort to obtain sufficient quantities of bio-oil for corrosion testing as well as for thorough chemical analysis, fractions of biomass-derived oils were acquired from Iowa State University (ISU). Two biomass sources were used for production of these bio-oils: red oak and corn stover. The ISU pyrolysis system separates the biomass-derived products into several different boiling point fractions. These various fractions have been analyzed and corrosion tests conducted with the same fractions (when sufficient quantities were available). In addition, a transfer line from the pyrolysis system has been destructively examined and corrosion samples are being exposed in the freeboard of the pyrolyzer. This paper describes some of the analytical techniques and analysis results, as well as the results of the laboratory corrosion studies. The results from examination of the transfer line, as well as a component from another liquefaction system, will be described.

Bio-oil CharacterizationBio-oil produced at ISU from fast pyrolysis of red oak and corn stover was chemically analyzed by Gas Chromatography/ Mass Spectroscopy (GC/MS). Additional details on the nature and processing steps of the supplied samples have been previously reported.7 The column was an HP-5MS (5% Phenyl)-methylpolysiloxane column, 0.25mm inner diameter, 0.25 m film thickness, and 30m length. The carrier gas was ultra high purity helium (99.9995%) with a constant flow rate of 1.0mL/min. The oven temperature was programmed at 30C for 2 minutes, followed by a temperature ramp to 200C at a rate of 10/minute, ending with a temperature hold at 200C for 10 minutes. A filtered, neat liquid sample with a volume of 1 L was injected onto the column with a split ratio of 1:100 at an injector temperature of 250C. The MS was set to scan the m/z range from 10 to 300 operating in the electron ionization mode. Peak identification was accomplished using the GCMS Chemstation Data Analysis software and by comparing mass spectra to the National Institute of Standards and Technology 08 mass spectral library.Borrowing from the petroleum energy industry, new approaches were developed for accommodating polarity and oxygenation. In order to glean meaningful quantification of total acid number (TAN) from many pyrolysis oil sample matrices, which are immiscible with the nonpolar carrier solvent proscribed in ASTM D664, modifications were applied to that method, henceforth referred to as ModTAN. These modifications include the use of high sonic energy aqueous extraction techniques and preparation of a hydrophilic titration solvent. In order to speciate the compounds, specifically the organic acids, present in the bio-oils, two parallel sample preparations and subsequent analyses are used. First, capillary electrophoresis (CE) in an indirect ultraviolet detection mode is used to separate the anions (conjugate bases to the organic acids) produced at high pH during the ModTAN titrations. In parallel, acidified samples are extracted in a non-polar solvent system for separation of the now-neutral organic acids via gas chromatography with mass spectrometric detection in the electro-ionization sampling mode (GC/EI-MS). Finally, in exploratory work to identify some of the apparently larger molecular weight organic acids in the basic CE runs, that separation approach is coupled to a mass spectrometer in the chemical ionization mode (CE/CI-MS).Laboratory Corrosion StudiesLaboratory corrosion studies have been conducted and are continuing with red oak and corn stover-derived pyrolysis oils to evaluate the corrosion behavior of candidate structural materials. The majority of the laboratory studies have been conducted with as-produced bio-oil that was exposed at 50C. The 50C test temperature was selected because that is likely to be the maximum temperature at which as-produced bio-oil would be transferred or stored, and because the bio-oils have limited stability at higher temperatures. The tests at 50C with the bio-oil consist of exposing two types of samples of five different alloys for a total of 1,000 hours with the samples being removed from the oil and examined after 250, 500 and 1,000 hours of exposure. Both general corrosion coupons and U-bend stress corrosion cracking coupons of carbon steel, 2Cr-1Mo steel, 409 stainless steel, 304L stainless steel and 316L stainless steel are used in all tests. Nominal compositions of these alloys are given in Table 1. Table 1. Nominal Compositions of Alloys Used for Corrosion StudiesAlloy Fe (wt%) Cr (wt%) Ni (wt%) Mo (wt%) Mn (wt%) C (wt%)

Carbon steel Balance ---- ---- ---- 1.0 0.13

2Cr-1Mo steel Balance 2.25 ---- 1.0 0.4 0.1

409 stainless steel Balance 11 ---- ---- 0.3 0.015

304L stainless steel Balance 18.3 9.0 ---- 1.7 0.02

316L stainless steel Balance 16.4 10.2 2.1 1.6 0.02

As described in three previous publications,3-5 one set of samples is immersed in the bio-oil and a second set is exposed in the vapor space above the oil. A water-cooled condensing column on the test system and a very slow flow of argon cover gas are used to prevent loss of very much, if any, volatile material. Pictures of two of the test systems and typical samples are shown in Figure 1.

Figure 1. Pictures of two of the test systems and typical samples Because these systems require about 2 liters of bio-oil for each test, the need for this quantity of liquid has prevented testing of some of the fractions because insufficient quantities were produced. Therefore, only chemical characterization studies could be conducted.Sample and Component ExaminationsSamples exposed in the corrosion tests as well as samples exposed in operating systems and components of such systems were thoroughly examined to determine the extent of degradation, and if possible, the cause of the degradation. The U-bend samples exposed in the laboratory corrosion tests as well any component in which cracking is a concern are subjected to a dye penetrant inspection to look for any evidence of cracking on the side of the samples subjected to a tensile stress. If a significant corrosion scale is present on a sample, the scale is collected and examined using X-ray diffraction. This technique was used to identify iron formate as the compound present on the surface of a few carbon steel and 2Cr-1Mo steel samples as was reported in previous publications.4-6 Following these nondestructive studies, cross sections are generally cut from the metallic samples for examination of the surface condition and the subsurface microstructure with the light microscope to help determine the extent and mechanism of reaction with the environment. If more sophisticated tools are needed, these sections are examined with the scanning electron microscope and if a more detailed elemental analysis is needed, the electron microprobe would be employed.RESULTSBio-oil CharacterizationResults of measurement of the acidity (ModTAN) and the concentration of carboxylic acids for the fractions produced by pyrolysis of red oak and corn stover are shown in Table 2. Table 2. Results of ModTAN and Acid Concentration Measurements on Bio-oil Produced by Pyrolysis of Red Oak and Corn StoverModTAN Formate content (ppm) Acetate content (ppm)

Red Oak fraction 1 131.6 299.3 9603.5

Red Oak fraction 2 70.6 2062.4 4857.4

Red Oak fraction 3 149.9 1609.5 20567.2

Red Oak fraction 4 143.7 2568.4 33231.7

Red Oak fraction 5 97.9 Not Available Not Available

Corn Stover fraction 1 47.0 515.9 4384.9

Corn Stover fraction 2 78.3 503.1 7546.6

Corn Stover fraction 3 133.7 1834.5 16749.5

Corn Stover fraction 4 124.6 Not Available 10607.9

Corn Stover fraction 5 79.3 Not Available 13318.8

These results show that fractions 3 and 4 (the lower boiling point fractions) have relatively high acidity, and this can be attributed to the high concentrations of formic and acetic acids.For the recovery method used in the ISU pyrolysis system, bio-oil is recovered as a series of stage fractions (SF) with distinct chemical properties using a system of condensers with controlled coolant temperatures and electrostatic precipitators. Samples included 5 fractions identified as SF1 to SF5.Stage fractions 1 and 2 were tar like and nearly solid at room temperature. These fractions contain the higher molecular weight sugars (levoglucosan) and high molecular weight organic molecules and were deemed unsuitable to be analyzed by the GC/MS method described above. Stage fractions 3 and 4 were liquid at room temperature and had a pungent, smoky odor. Stage fraction 5 was primarily the aqueous phase and was highly acidic, but it could not be analyzed by GC/MS. To simplify the reporting of the compounds identified in the bio-oil, the compounds were separated into eight groups based on the distinguishing properties. These groups are: acids, aldehydes, alcohols, ketones, furans, phenols, monomethoxyphenols, and dimethoxyphenols.Figure 2 illustrates the relative concentration of each group of organic molecules present in stage fractions 3 and 4 for the red oak samples.

Figure 2. Relative concentration of organic compounds identified in red oak bio-oil stage fractions determined by GC/MS. Figure 3 illustrates the relative concentration of each type of organic molecules present in stage fractions 3 and 4 for the corn stover samples. The bio-oil derived from red oak feedstock appeared to be more acidic than the corn stover and would be expected to be slightly more corrosive to metallic components. The red oak bio-oil has a fairly even distribution of compounds amongst the different organic groupings. The bio-oil derived from the corn stover had a fairly even distribution of compounds among the grouping with the exception of having an elevated phenolic content in both stage fractions 3 and 4. There does not appear to be a significant difference between the chemical content of the stage fractions 3 and 4 for both the red oak and corn stover samples.

Figure 3. Relative concentration of organic compounds identified in corn stover bio-oil stage fractions determined by GC/MS.Laboratory Corrosion StudiesThe bio-oils whose characterization was described in the previous section were also used in corrosion tests to obtain some indication of their relative corrosion behavior. Because the available quantity of some of the oils was limited, some fractions had to be combined to have a sufficient amount of oil to conduct the studies. Results of the corrosion rate calculations based on weight change of the corrosion samples are shown in Table 3.Table 3. Calculated Corrosion Rates (mm/year) after 500 Hour Exposure for Samples Exposed in Bio-oil Produced from Fast Pyrolysis of Red Oak and Corn StoverBio-Oil Carbon steel 2Cr-1Mo steel 409 stainless 304L stainless 316L stainless

Coupon U-bend Coupon U-bend Coupon U-bend Coupon U-bend Coupon U-bend

Immersed in oil

Red Oak #3 & #4 2.96 2.90 2.45 2.61 0.44 0.90