*Article* **A Detailed Database of the Chemical Properties and Methane Potential of Biomasses Covering a Large Range of Common Agricultural Biogas Plant Feedstocks**

**Audrey Lallement 1, Christine Peyrelasse 1, Camille Lagnet 1, Abdellatif Barakat 2, Blandine Schraauwers 1, Samuel Maunas <sup>1</sup> and Florian Monlau 1,\***


**Abstract:** Agricultural biogas plants are increasingly being used in Europe as an alternative source of energy. To optimize the sizing and operation of existing or future biogas plants, a better knowledge of different feedstocks is needed. Our aim is to characterize 132 common agricultural feedstocks in terms of their chemical composition (proteins, fibers, elemental analysis, etc.) and biochemical methane potential shared in five families: agro-industrial products, silage and energy crops, lignocellulosic biomass, manure, and slurries. Among the families investigated, manures and slurries exhibited the highest ash and protein contents (10.3–13.7% DM). High variabilities in C/N were observed among the various families (19.5% DM for slurries and 131.7% DM for lignocellulosic biomass). Methane potentials have been reported to range from 63 Nm<sup>3</sup> CH4/t VS (green waste) to 551 Nm3 CH4/t VS (duck slurry), with a mean value of 284 Nm3 CH4/t VS. In terms of biodegradability, lower values of 52% and 57% were reported for lignocelluloses biomasses and manures, respectively, due to their high fiber content, especially lignin. By contrast, animal slurries, silage, and energy crops exhibited a higher biodegradability of 70%. This database will be useful for project owners during the pre-study phases and during the operation of future agricultural biogas plants.

**Keywords:** anaerobic digestion; agricultural inputs; biochemical methane potential; biodegradability; lignocellulosic biomasses; manures

#### **1. Introduction**

Biogas production has increased in the European Union, encouraged by the European "Green Deal" and the renewable energy policies [1,2]. Between 2000 and 2017, global biogas production quadrupled, from 78 to 364 TW h, which corresponds to a global yearly volume of 61 billion m3 biogas; it is shared mainly among Europe (54%), Asia (31%), and the Americas (14%) [1]. Anaerobic digestion (AD) unit numbers are increasing in Europe, supported by the need to improve green energy supplies. Among the typologies of biogas plants, agricultural biogas plants are gaining increasing interest as a valuable technology to treat agricultural residues and co-products, thereby generating energy and fertilizers and improving farmers' incomes. In 2021, France had approximately 401 AD on farms and 285 centralized or territorial AD (Source: SINOE). In parallel, in 2018, 1555 and 9500 biogas plants were reported in Italy and Germany, respectively [1]. Nonetheless, it appears that the biogas sector is facing a shift in its development paradigm [1]. At the European level, the biogas sector is still dominated mainly by a model based on energy crops, high feed-in tariffs, and local electrical production via combined heat and power units. However, the biogas sector is now moving towards a different model, where organic wastes, agricultural by-products, as well as sequential crops are used mainly as feedstocks, and biogas is upgraded to biomethane for various applications (transportation, chemical production, heat, etc.) [1].

**Citation:** Lallement, A.; Peyrelasse, C.; Lagnet, C.; Barakat, A.; Schraauwers, B.; Maunas, S.; Monlau, F. A Detailed Database of the Chemical Properties and Methane Potential of Biomasses Covering a Large Range of Common Agricultural Biogas Plant Feedstocks. *Waste* **2023**, *1*, 195–227. https:// doi.org/10.3390/waste1010014

Academic Editors: Dimitris P. Makris and Vassilis Athanasiadis

Received: 6 November 2022 Revised: 20 December 2022 Accepted: 22 December 2022 Published: 10 January 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

As the number of biogas plants has increased, securing deposits and the need for alternative feedstocks are growing. The main families of inputs for agricultural biogas plants are animal wastes (manures and slurries), lignocellulosic biomasses, energy and sequential crops and silages, and agricultural co-products. To help industrial and biogas operators, a better knowledge of the main chemical properties (organic matter, fibers, proteins, elemental analysis, C/N, COD, etc.) along with biochemical methane potential tests are needed. The C/N ratio of feedstock is another important parameter, and for a good anaerobic digestion process, the C/N ratio must be between 20 and 30 [3,4]. Indeed, if a biogas reactor has a low C/N ratio, there is potential inhibition from ammonia [3,5]. Among the chemical parameters, the content of fibers (cellulose, hemicelluloses, and lignin) and proteins is another important issue that can affect the final biodegradability of substrates [6]. Finally, the information of the elemental analysis (C, H, N, S, and O) is of prime importance, as it will allow determination of both the theoretical chemical oxygen demand (COD) and the theoretical methane potential according to the Buswell equation [7].

Aside from chemical properties, the determination of the methane potential through BMP (biochemical methane potential) tests is important. BMPs allow laboratory-scale measurement of the maximum production of methane generated by the digestion of a single substrate, and in recent decades, several national and international inter-laboratory studies have been carried out to optimize the protocol and define good practices [8–10]. BMP tests are a popular technique to determine the methane potential and biodegradability of organic substrates [11]. Currently, the BMP test is used for the technical and economic analysis of a project, for the design of agricultural biogas plants, and for evaluation of the process performance [8]. BMP tests can also be useful when the biogas plant unit is operating and new biomasses are to be introduced. Table 1 lists recent studies that provided detailed BMP data of different organic wastes along with the ISR (inoculum-to-substrate ratio) applied. Indeed, the ISR of the BMP is one of the crucial parameters, and the generally recommended values are between 2 and 4 [9,11]. In parallel with classical BMP tests, a theoretical one can also be estimated according to the elemental composition and the Buswell equation [12] or the COD [13], the chemical composition (lipids, carbohydrates, and proteins) [13], or using McCarty's method [14], allowing determination of the biodegradation rate of a selected substrate.

It is of interest to note that a few publications have provided detailed methane potentials per substrate categories, and they generally provide only min., max., and mean values. Among these publications, Allen et al. (2016) reported methane potentials for 83 organic substrates covering different categories from first-, second-, and third-generation biomasses with agricultural wastes, agro-industrial wastes, food residues, and seaweeds [5]. In parallel, Garcia et al. (2019) reported a detailed methane potential database of more than 50 agricultural and food processing substrates [15]. Similarly, Godin et al. (2015) referenced the methane potential of 569 plant biomasses [16]. In parallel, other studies reported exhaustive lists of the methane potentials of 56 agricultural wastes [17], 48 maize sample silages [18], 43 crop species [19], 12 lignocellulosic biomasses [6], and 30 organic wastes [14].

To date, there is clearly a lack of information in the literature regarding data about the chemical and methane potentials of a large spectrum of agricultural biogas plant feedstocks. This publication aims to highlight the characterization of 132 substrates shared by five different families: cereal and residue (CER), energy crop and silage (ENSI), lignocellulosic matter (LCM), manure (MAN), and slurry (SLU). The selection of substrates was based on their frequency of inclusion in agricultural biogas plants. First, various chemical properties (organic matter, fibers, proteins, elemental analysis, C/N, COD, etc.) were analyzed for the 131 substrates. Then, methane potentials were assessed on these substrates and biodegradability rates (defined as the ratio of the BMP assay yield to the theoretical Buswell yield) were calculated.

**Table 1.** Literature data on large sets of BMP references for organic substrates. N.: number of samples, ISR: inoculum–substrate ratio, MSW: municipal solid wastes, and WWTP: wastewater treatment plant. Description of the samples can be found in the Appendix B.



\* data are not provided directly in the publication but in an appendix of the authors publications.

#### **2. Materials and Methods**

#### *2.1. Sampling*

Feedstocks were collected in thirty agricultural biogas plant units operating with agricultural feedstocks on the national level. Of these, 75% were operating in wet AD and 25% in dry AD. These 132 inputs are regrouped into five main families: cereal and agro-industrial co-products (CER), energy crop and silage (ENSI), lignocellulosic matter (LCM), manure (MAN), and slurry (SLU). A description of the dataset is available in the Appendix A (Tables A1 and A2).

#### *2.2. Elemental Composition and Fiber Analysis*

The elemental composition of each feedstock was assessed by an elemental apparatus (varioMicro V4.0.2, Elementar®, Langenselbold, Germany), after being dried at 60 ◦C until constant weight and ground into 1 mm particles using a centrifuge mill (SR 200, Retsch, Haan, Germany). Each COD was then calculated on the basis of this analysis using Equation (1) [31]:

$$\text{COD} \left( \frac{\text{gCOD}}{\text{gCxHyOz}} \right) = 8 \times \frac{4\pi + y - 2z}{12\pi + 4 + 16z} \tag{1}$$

The protein content was estimated on the basis of the nitrogen elemental composition multiplied by 6.25 [32].

For fiber analysis (e.g., cellulose, hemicelluloses, and lignin-like), 80 mg of sample was hydrolyzed with 0.85 mL of H2SO4 acid (72%) for 1 h at 30 ◦C in continuously shaken tubes for thorough mixing (450 rpm) using closed vessels to prevent evaporation. Then, 23.8 mL of deionized water was added, and the vessels were heated to 121 ◦C for one hour under magnetic agitation (450 rpm). After cooling, the insoluble residue was separated by filtration through 1 μm glass fiber paper (GFF, WHATMAN®, Maldstone, UK) into a soluble phase (structural carbohydrates) and a solid phase (lignin and ash). The filtrate was further filtered using nylon filters (0.2 μm) and analyzed for glucose, xylose, and arabinose by high-performance liquid chromatography (1260 infinity II technology, Agilent, Santa Clara, CA, USA) equipped with a Hi.Plex H coupled to a UV detector. The crucible and the fiberglass paper were dried at 105 ◦C for 24 h to determine the content of Klason lignin-like

material by weighing. The cellulose-like and hemicelluloses-like contents were determined using the following equations:

$$\text{Cellulose} - \text{like } (\% \text{ DM}) = \frac{\text{Glucose } (\% \text{ DM})}{1.11} \tag{2}$$

$$\text{Hemicelluloses} - \text{like } (\% \text{ DM}) = \frac{\text{Xylose } (\% \text{ DM}) + \text{Arabanose } (\% \text{ DM})}{1.13} \tag{3}$$

where 1.11 is the conversion factor of polymers based on glucose-to-glucose monomers, and 1.13 is the factor for converting polymers based on xylose (arabinose and xylose) into monomers [33].

#### *2.3. Biochemical Methane Potential Measurement (BMPexp)*

The procedure for BMP tests has been well-documented in a previous study [30] and followed the inter-laboratory study recommendations [8,34]. Feedstocks were stored at 5 ◦C if the storage period was less than or equal to three days or at −20 ◦C if the storage period exceeded three days and thawed at 6 ◦C before testing. Used inoculum was agitated, maintained at 38 ± 1 ◦C, and fed regularly with green grass and wastewater sludge at the laboratory of APESA facility. Regular checks were performed by measuring the pH, dry matter, and volatile solids. DM and vs. were obtained by loss on ignition (same as for feedstocks), and the pH was assessed using a 340i pH meter fitted with Sentix® electrodes (WTW, Weilheim, Germany). The main properties of the inoculum were TS (% fresh mass): 3.8 ± 0.3%; vs. (% TS): 64.4 ± 1.5%; pH: 8.3 ± 0.2; volatile fatty acids (VFAs): 300 mg eq. acetate L−1; and ammonium content: 2.1 g N-NH4 <sup>+</sup> L−1. The inocula complied with the quality criteria proposed by [10].

The BMP tests were carried out under mesophilic conditions in duplicate, and 500 mL reactors were filled with 300 mL of an inoculum/substrate ratio of 3 g VS/g VS. After filling, each bottle was flushed with N2 gas for 30 s, incubated at 39 ◦C, and degassed after 1 h. Each day, manual homogenization was performed, and biogas production followed using an electronic manometer device (Digitron 2023P, Digital Instrumentation Ltd., London, UK) and expressed in normal liters (at 0 ◦C, 1.013 hPa). Once a week, the gas composition was analyzed by gas chromatography (Varian GC-CP4900, Agilent, Santa Clara, CA, USA) equipped with two columns. For O2, N2, and CH4, a Molsieve 5A PLOT column at 110 ◦C was used, and for CO2 analysis, a HayeSep A set at 70 ◦C was used. The injector and detector temperatures were set at 110 ◦C and 55 ◦C, respectively. Two standard gases for calibration were used: one composed of 9.5% CO2, 0.5% O2, 81% N2, and 10% CH4, and the other composed of 35% CO2, 5% O2, 20% N2, and 40% CH4 (special gas from Air Liquide®, Paris, France). The BMP tests concluded when the biogas production reached a stationary state and did not vary for more than 0.5% during three consecutive days. Blank (inoculum only) and positive controls (cellulose, Tembec®, Montréal, QC, Canada) were run in parallel in duplicate.

The theoretical BMP was calculated on the basis of the elemental characterization (CxHyOzNnSs) using Equation (4) (Achinas and Euverink, 2016):

$$BMPth\left(LCH\_4/kg\ VS\right) = \frac{22.4 \times \left(\frac{x}{2} + \frac{y}{8} - \frac{z}{4} - \frac{3u}{8} - \frac{s}{4}\right)}{12x + y + 16z + 14u + 32s} \tag{4}$$

where 22.4 is the molar volume of an ideal gas.

Finally, the percentage of biodegradation is the ratio between the experimental BMP and the theoretical BMP.

$$Bioderadiation\ \left(\%\right) = \frac{BMPexp}{BMPth} \tag{5}$$

#### **3. Results**

#### *3.1. Chemical Composition of the Various Biomasses*

The feedstock compositions are described in Figure 1 (overall and for each family, more data are available in SD Table 1). Among the five families, the distribution was as follows: 47% energy crops and silages, 32% lignocellulosic biomasses, 31% manures, 17% cereal co-products and residues, and 5% slurries. The minimum, maximum, and average values of the different chemical properties (DM, VS, C/N, fibers, proteins, and COD) are shown for all the families in Tables A1 and A2. In order to have a better sense of the inter-family variability, the most important parameters (i.e., VS/DM, COD, C/N, and protein content) are presented as boxplots (Figure 1) and the fiber compositions as radar graphs (Figure 2).

**Figure 1.** (**A**–**D**) Boxplots of chemical composition variabilities: VS/DM, COD, C/N, and protein content. Medians are the horizontal lines and means are represented by squares. Families: cereal and residue (CER), energy crop and silage (ENSI), lignocellulosic matter (LCM), manure (MAN), and slurry (SLU).

First of all, higher ash contents were reported for manures and slurries compared with the other families investigated. In terms of proteins, higher contents were also reported for manures and slurries. Indeed, mean protein contents of 10.4 and 13.7% DM were reported for animal manures and slurries, respectively. By contrast, lignocellulosic biomasses exhibited the lowest protein content, at 3.9% DM. Allen et al. (2016) reported protein contents varying from 12.3% DM to 18.5% DM for different animal slurries [5]. Similarly, Li et al. (2013) estimated protein contents of 13.7% DM, 17.5% DM, and 21% DM for swine, dairy, and chicken manures, respectively. Li et al. (2013), on the other hand, reported lower values ranging from 2.5% DM to 5.6% DM for lignocellulosic biomasses [24]. The chemical oxygen demand is another important parameter in anaerobic digestion monitoring, as it can allow determination of mass balances and the theoretical methane potential [13]. Little variability in the main COD was observed for the five families, with values ranging from 1.3 to 1.5 g/g VS. Scarce information is available in the literature regarding these parameters, as only Labatut et al. (2011) have reported it for a range of 30 substrates (monoand co-digestion). For manure, they found a COD ranging from 0.7 to 1.3 g/g, with a mean of 1.0, which is considerably lower than our values, and higher values for biowaste substrates, with a mean of 6.4, ranging from 0.9 to 28.8 g/g [14]. The fiber content (i.e., cellulose, hemicelluloses, and lignin) was also reported for the five families, and higher contents were observed for lignocellulosic biomasses and manures, similar to the values previously reported in the literature [6,14,24].

Finally, the C/N ratio was also reported for the five families. The C/N ratio is a very important parameter for the long-term continuous digestibility of a substrate. Ideally, it should be between 25:1 and 30:1 to facilitate optimal growth of micro-organisms [5]. For this parameter, high variabilities were observed with higher values of C/N for lignocellulosic biomasses, with a median of approximately 90 and an average of 132. All the other groups have means or averages between 19 and 40. Yet, the C/N ratio is based on the elemental analysis, requiring dry samples. Volatilization of ammoniacal nitrogen or volatile compounds can differ depending on the substrate. A comparison of these results with C/N ratios in the literature points out that an overestimation occurred for slurry and manure families [15,35,36]; similar results are obtained for CER and LCM [5,15], whereas ENSI family C/N ratios are underestimated [5,15,37,38]. Extrapolations cannot be readily performed, as they can depend on the feedstock composition, type, harvest, storage, etc. As an example, manure C/N ratio means have been found to be approximately 16 for cattle manure, 9 for poultry manure, and they are higher for horse manure (between 15 to 150, depending on the type and proportion of litter) [35,36,39].

**Figure 2.** Means of the different family composition of fibers. Families: cereal and residue (CER), energy crop and silage (ENSI), lignocellulosic matter (LCM), manure (MAN), and slurry (SLU).

#### *3.2. Biochemical Methane Potential of Feedstock*

Another important parameter in the monitoring and optimization of agricultural biogas plants is the value of the methane potential. Methane potentials were assessed in this study by BMP tests performed on the 132 agricultural substrates shared in five families: cereals and agro-industrial co-products, lignocellulosic biomass, energy crops and silages, animal manures, and slurries (Figure 3).

**Figure 3.** Methane potentials of the different categories of the five families. Families: cereal and agro-industrial residue in grey (CER), energy crop and silage in blue (ENSI), lignocellulosic matter in green (LCM), manure in red (MAN), and slurry in yellow (SLU).

As shown in Table 1, a large variability in methane potentials was observed among the different families, with methane potentials ranging from 63 Nm<sup>3</sup> CH4/t VS (green waste) to 551 Nm<sup>3</sup> CH4/t VS (duck slurry), with a mean value for the 132 organic samples of 284 Nm3 CH4/t VS.

#### 3.2.1. Cereal and Agro-Industrial Residues (CER)

The first family investigated was cereal and agro-industrial residues (*n* = 17). The cereals were obtained from the cereal agro-industry and silos, whereas the maize was

from the sweet corn industry. Methane potentials of 298, 301, and 318 Nm3 CH4/t VS were reported for cereal residues, sweet corn residues, and wheat residues, respectively. Garcia et al. (2019) reported a similar methane potential, with values of 345 Nm3 CH4/t VS for a mix of cereals [15]. Luna DeRisco et al. (2011) also investigated the methane potentials of grain mill residues, and methane potentials of 274–386 Nm3 CH4/t VS were reported [40]. In parallel, Garcia et al. (2019) also reported methane potentials ranging from 204 to 345 Nm3 CH4/t VS for ten agro-industrial co-products (from the vegetables and fruits industry) [15].

#### 3.2.2. Manures (MAN)

The methane potential of various animal manures was investigated. Manures are organic matter, derived mostly from animal feces and urine but also normally containing plant materials (generally wheat straw) that have been used as bedding for animals. Methane potentials of 173, 210, 217, 230, 235, and 250 Nm3 CH4/t VS were determined for turkey, cattle, pig, poultry, zoo, and horse manures. Such data are in the same range as the values reported in the literature [24,41,42]. Kafle and Chen (2016) investigated the methane potential of five different livestock manures (dairy manure (DM), horse manure (HM), goat manure (GM), chicken manure (CM), and swine manure (SM)). The BMPs of DM, HM, GM, CM, and SM were determined to be 204, 155, 159, 259, and 323 Nm<sup>3</sup> CH4/t VS, respectively [41]. Similarly, Cu et al. (2015) also reported methane potentials of various animal manures, and the highest BMP in this study was from piglet manure at 443.6 Nm3 CH4/t VS, followed by cow, sow, chicken, rabbit, buffalo, and sheep manures at 222, 177.7, 173, 172.8, 153, and 150.5 Nm3 CH4/t VS, respectively [42]. Similarly, Garcia et al. (2019) reported methane potentials of 97, 128, 200, and 208 Nm3 CH4/t VS for bovine, pig, rabbit, and poultry manures, respectively [15]. Yang et al., 2021 also reported methane potentials of 160 Nm<sup>3</sup> CH4/t VS for dairy manure, 200 Nm3 CH4/t VS for goat manure, and 325 Nm<sup>3</sup> CH4/t VS for swine manure [43]. It can be observed that the methane potentials of our studies are in the same range as the literature data, although some differences can be observed for the same manure families, as the methane potential can be influenced by the type of farm, the duration of storage, and the storage method. Finally, Carabeo-Perez et al. (2021) also investigated the methane potential from various herbivorous animal manures. Methane yield potentials of 245, 326, and 112 Nm<sup>3</sup> CH4/t VS were obtained for horse, rabbit, and goat manures, respectively, influenced by the difference in their digestive systems to digest the grass feedstock [44]. Finally, Li et al. (2013) determined methane potentials of 51, 295, and 321 Nm<sup>3</sup> CH4/t VS for dairy, chicken, and swine manure, respectively [24].

#### 3.2.3. Animal Slurries (SLU)

Animal slurries are manure in liquid form, i.e., a mixture of excrements and urine of domestic animals, including water and/or small amounts of litter. Slurry methane potentials were also investigated in this study, with methane potentials ranging from 263 to 551 Nm<sup>3</sup> CH4/t VS. As shown in Figure 3, a high variability was observed for cattle slurries, which can be explained by differences in the storage type and duration. In terms of liquid manures, little information is available in the literature [5,14]. Labatut et al. (2011) reported a methane potential of 261 Nm3 CH4/t VS for liquid dairy manure. Allen et al. (2016) investigated the methane potentials of different slurries (dairy, pig, and beef). Methane potentials of 99 and 311 Nm3 CH4/t VS were reported for pig and beef slurries, respectively. In terms of dairy slurries, methane potentials ranging from 136 to 239 have been reported [5]. Garcia et al. (2019) also reported methane potentials of 35 and 137 Nm3 CH4/t VS for bovine and pig slurries, respectively [15].

#### 3.2.4. Silages and Energy Crops (ENSI)

Silages and energy crops are another type of substrate generally found in agricultural biogas plants. In our study, of the 46 organic substrates investigated, the methane potentials ranged from 187 Nm3 CH4/t VS to 461 Nm<sup>3</sup> CH4/t VS. For instance, average methane

potentials of 320, 342, and 352 Nm3 CH4/t VS were reported for sorghum, corn, and grass samples, respectively. The methane potentials of silages and energy crops have been widely investigated in the literature in recent decades, and the values obtained in this study are in the same order [5,15,18,19]. For instance, Garcia et al. (2019) investigated the methane potential of five energy crops and reported methane potentials ranging from 253 Nm3 CH4/t VS (millet, *Panicum milliaceum* L.) to 351 Nm<sup>3</sup> CH4/t VS (triticale, *Triticum aestivum* L.). Similarly, Allen et al. (2016) reported the methane potential of 18 energy crops, and the methane potentials ranged from 281 Nm3 CH4/t VS (winter oats) to 398 Nm<sup>3</sup> CH4/t VS (turnips). Similarly, Allen et al. (2016) also investigated the methane potentials of different silages and reported methane potentials varying from 311 Nm3 CH4/t VS (Savazi grass silage) to 433 Nm<sup>3</sup> CH4/t VS (silage bales). Finally, Hermann et al. also investigated the methane potentials of 43 crops, including main and secondary crops, catch crops, annual grass, and perennial crops [19].

#### 3.2.5. Lignocellulosic Biomasses (LCM)

The methane potentials of 33 lignocellulosic biomasses were also investigated. The methane potentials ranged from 63 Nm3 CH4/t VS (green waste) to 330 Nm<sup>3</sup> CH4/t VS (barley straw). Lower methane potentials were observed for green waste residues, likely due to their high content in fibers, and especially in lignin, which has been shown to be poorly degraded in the anaerobic digestion process [19,45]. Similar methane potentials on lignocellulosic biomasses have been reported previously in the literature [6,15]. Indeed, Monlau et al. (2012) reported the methane potentials of twelve lignocellulosic biomasses ranging from 155 Nm3 CH4/t VS (sunflower stalks) to 300 Nm<sup>3</sup> CH4/t VS (Jerusalem artichoke tubers). Similarly, Garcia et al. (2019) reported methane potentials ranging from 282 Nm<sup>3</sup> CH4/t VS (coconut fibers) to 425 Nm<sup>3</sup> CH4/t VS (corn, *Zea mays* L.). Similarly, Dinuccio et al. (2010) reported methane potentials ranging from 225 to 424 Nm3 CH4/t VS [46]. Perennial crops exhibited the lowest methane potentials, with values ranging from 203 Nm3 CH4/t VS (cup plant) to 260 Nm3 CH4/t VS (tall wheatgrass). The highest methane potential of the various crops investigated was reported for forage triticale, with a methane potential of 371 Nm<sup>3</sup> CH4/t VS.

#### *3.3. Practical Implementation of this Database*

To assist the reader and user in exploiting this publication, a summary table is provided in Table 2 with the main physicochemical parameter and methane potential values for the various substrate families investigated in this study. As previously discussed, the methane potentials ranged from 63 Nm<sup>3</sup> CH4/t VS (green waste) to 551 Nm3 CH4/t VS (duck slurry), with a mean value for the 132 organic samples of 284 Nm3 CH4/t VS.

To better understand the ability of the various organic wastes that were tested to be degraded in the AD process, a biodegradation yield (based on the ratio of the experimental and theoretical BMP) was calculated using the Buswell formula. The family biodegradation yields are presented in Figure 4.


**Table 2.** Chemical composition of the families (FM: fresh matter; DM: dry matter; and VS: volatile solids). Families: cereals and residues (CER), energy crops and silage (ENSI), lignocellulosic matter (LCM), manures (MAN), and slurries (SLU).

**Figure 4.** Boxplots of biodegradation yields of the five families. Medians are the horizontal lines and means are represented by squares. Families: cereal and residue (CER), energy crops and silage (ENSI), lignocellulosic matter (LCM), manure (MAN), and slurry (SLU).

A majority of families presented a good biodegradation rate, with means between 52 and 73%. Lower degradation rates of only 52 and 56% were reported for manure and lignocellulosic matter, respectively. As manure is a mixture of feces and bedding material, depending on the bedding material used and its concentration, it is not surprising to find similar results between these two families [47]. The biodegradability of organic substrates has been well-documented in the literature for various organic substrates [5,14,15,17,24]. Regarding lignocellulosic biomasses, Triolo et al. (2012) reported biodegradability indices of 32.7%, 39.9%, 44.9%, and 66.6% for wood cuttings, hedge cuttings, wild plants, and lawn cuttings, respectively. Similarly, Li et al. (2013) reported biodegradability indices of 51%, 54%, and 62% for corn stover, wheat straw, and rice straw, respectively. Similarly, Li et al. (2013) reported biodegradation rates of 10%, 63%, and 68% for dairy manure, chicken manure, and swine manure, respectively. Garcia et al. (2019) also reported biodegradability indices varying from 30% to 70% for different animal manures samples. Such lower biodegradation rates for LCM and MAN families can be explained by the higher fiber contents in such biomasses, especially lignin content, which is poorly degraded in the AD process [6,45]. The high nitrogen concentration in animal manures can also be a limiting factor of the expression of the methane potential [42].

In parallel, other families investigated in this study exhibited higher biodegradability rates of 69%, 72%, and 73% for cereal and agro-industrial residues, energy crops and silages, and slurries, respectively. Allen et al. (2016) reported biodegradability indices for sixteen silages from second-generation crops, and three-quarters of the samples exhibited biodegradation rates higher than 75%. Similarly, Garcia et al. (2019) reported biodegradabilities varying from 80% to 100% for various energy crops (i.e., millet, barley, maize, sorghum, and triticale). Garcia et al. (2019) also reported high biodegradabilities of 80% and 90% for flour and cereals. In terms of slurry samples, the results in the literature are more contrasted [5,15]. Indeed, biodegradabilities varying from 20% to 60% have been reported. Such variation can be explained by the difference in the origins of animal slurries as well as the storage duration and typology.

#### **4. Discussion**

At the end of 2018, annual production of biomethane from AD in the EU corresponded to 2.3 billion m3, with 18,202 biogas plants in operation [1]. Europe is the world leader in biogas electricity production, far ahead of the USA (2.4 GW) and China (0.6 GW) [1]. At the European level, the methanization sector will greatly develop in the years to come with projections up to 64.2 billion m<sup>3</sup> in the EU by 2050; this would represent an energetic potential of approximately 640 TW h/year and would require a 30-fold growth of the current biomethane sector [1].

AD will continue to grow in the future, but it is clear that the sector should have better control of not only the management and the use of the deposits but also the identification of new sources of deposit. The BMP test remains an essential tool for characterizing new deposits and determining their pricing.

This publication and the results (Table 1) are intended to contribute to providing data to the scientific community and biogas developers regarding the values of methane potentials and biodegradability indices of different organic substrates and complete previous studies on the subject (Table A3 in Appendix B). In parallel, this study is intended to be a tool for the sizing, optimization, and operation of the biogas sector. All the data obtained for the different feedstocks are available in the Appendix A.

It could be interesting in the future to extend this work and to generate an overall synthesis of all the BMP values listed in the literature by taking into account the studies using a protocol based on the recommendations of interlaboratory guidelines carried out at the international level [10,34]. In parallel, the growing development of the biogas sector requires the mobilization of new resources and organic biomasses, and it will be interesting in the future to focus studies on the evaluation of the methanogenic potential of atypical biomasses (i.e., algae, paper sludges, biodegradable plastics, insect excrements, etc.). An

extended open-source BMP database (based on BMP values validated by experts) could be very useful in the future in order to improve the biogas development as well as the monitoring of the energetic performances of biogas plants. Indeed, Holliger et al. (2017) compared methane production from BMPs with biogas production from the same organic materials in full-scale installations [48]. Holliger et al. (2017) highlighted that the measured weekly methane production accounted for 94.0 ± 6.8 and 89.3 ± 5.7% of the calculated weekly methane production for two biogas plants, respectively [48].

Short-term (i.e., 1–2 months), batch-mode anaerobic digestion tests, such as the biochemical methane potential (BMP) assay, are intended primarily to determine methane yields and the biodegradability of substrates [14]. Nonetheless, such testing may fail to truly predict the performance of full-scale anaerobic reactors. For this purpose, semi-continuous laboratory-scale experimental methods are complementary to chemical and BMP analysis. Semi-continuous flow reactors are designed to emulate the conditions of commercial-scale digesters and study their overall performance over time, taking into account co-digestion benefits and potential inhibition.

#### **5. Conclusions**

In this study, a characterization of 132 common agricultural feedstocks (shared in five families) was carried out in terms of physical properties and methane potentials. Of the various families investigated, manures and slurries exhibited the highest ash and protein contents (10.3–13.7% DM). A high degree of variability in terms of the C/N ratio was observed among the various families, with values ranging from 19.5% DM (slurries) to 131.7% DM (lignocellulosic biomass). In terms of biodegradability, lower values of 52% and 57% were reported for lignocelluloses biomasses, and manures due to their high content in fibers, especially lignin. The AD sector will continue to grow in the future, and such studies can be used as a reference for any operator/manager of units or public authority/financial provider in the future.

**Author Contributions:** A.L., validation, investigation, and writing—original draft; C.P., investigation and writing—original draft; C.L., supervision and investigation; A.B., investigation and writing review and editing; B.S., methodology, analysis, and investigation; S.M., methodology, analysis, and investigation; F.M., financial support, supervision, conceptualization, and writing—review and editing. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work is incorporated in the SPIRALE project funded by the ADEME (GRAINE 2018; 1806C0002).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** All the data are described in Figures and Tables or in the appendix.

**Acknowledgments:** This work is incorporated in the SPIRALE project funded by the ADEME (GRAINE 2018; 1806C0002), whom we thank for their support along with the two other partner projects: Green Tropism and INSA Toulouse. The APESA also thanks the various operators who provided the biomasses used in this study.

**Conflicts of Interest:** The authors declare no conflict of interest.


 **A** **Table A1.** Description of the substrates analyzed within the families, where SD is standard deviation, DM is dry matter, FM is fresh matter, VS isvolatilesolids,BMPexpistheBMPmeasured,andBMPisthemaximummethanepotentialbasedonCHNScomposition.



 —







is





[18]

 204

 295

 329

 355


*Waste* **2023**, *1*





*Waste* **2023**, *1*



#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

## *Article* **Washing Methods for Remove Sodium Chloride from Oyster Shell Waste: A Comparative Study**

**Jung Eun Park, Sang Eun Lee and Seokhwi Kim \***

Center for Bio Resource Recycling, Institute for Advanced Engineering, Yongin-si 17180, Republic of Korea **\*** Correspondence: shkim5526@iae.re.kr; Tel.: +82-31-330-7203

**Abstract:** The oyster shell is a valuable calcium resource; however, its application is limited by its high NaCl content. Therefore, to establish the use of oyster shells as a viable resource, conditional experiments were conducted to select optimum parameters for NaCl removal. For this purpose, we compared leaching methods with batch and sequential procedures, determined the volume of water used for washing, and evaluated the mixing speed. The batch system removed NaCl when washed for >24 h over a shell to water ratio of 1:5. Results from the batch experiments confirmed that washing twice can completely remove NaCl from the shells on a like-for-like basis. Additionally, the efficiency of washing was sequentially evaluated in terms of the number of washing cycles. Compared to batch experiments, continuous washing could remove NaCl in approximately 10 min at a shell to water ratio of 1:4. We found that regardless of the washing methods, the volume of water used for washing is key for enhancing NaCl removal. Consequently, increasing the volume of water used for washing coupled with a proper sorting of fine particles can help enhance the purity of calcium, which will enable the use of oyster shell as an alternate Ca-resource.

**Keywords:** oyster shell waste; washing method; sodium chloride; impurities; particle size

#### **1. Introduction**

Oysters are widely used as part of a healthy diet, as they help improve immunity and in the rapid absorption of zinc and iron; moreover, increased awareness of their health benefits has led to an increase in the consumption of oysters [1]. As of 2017, 639 tons per year of Pacific oysters are produced in Asia, which constitutes approximately 92% of the global oyster market [2]. An oyster comprises about 14% oyster meat and 86% shell; during oyster processing, the shells are disposed as waste into the sea or buried into the ground [3,4]. Shells buried into the ground pose problems, as they produce leachates and odors. Furthermore, the organic matter in the shells discarded into the sea may pollute the ocean ecosystem. Hence, there is a need for a rapid establishment of policy support to address the substantial increase in the number of shells generated at present and in the future. Moreover, it is desirable to utilize shells as resources without disposing them as waste, as they can contribute to an environmental and economic circular economy [5,6]. Both Korea and Southern Brazil have suggested recycling of shells for environmental benefits; as of 2018 [6,7], Japan has recycled 140,000 tons of oyster shells to improve the environment of fishing farms and to enhance the sediment quality by establishing underwater plants [8]. Taiwan generates 34,000 tons of oyster shells, whose disposal into the environment causes sanitation, pollution, and protection issues [9].

Oyster shells have been disposed as waste along the coast of Korea [10]; however, after the amendment of the Framework Act on Resources Circulation in 2022, the status of oyster shells has been changed from waste to resources to help control the influx of impurities such as organic residues. Nevertheless, there are two challenges associated with the application of oyster shells as an alternative to industrial material such as limestone: (i) the levels of impurities equivalent to those of limestone and (ii) the demand for a massive quantity [11–13].

**Citation:** Park, J.E.; Lee, S.E.; Kim, S. Washing Methods for Remove Sodium Chloride from Oyster Shell Waste: A Comparative Study. *Waste* **2023**, *1*, 166–175. https://doi.org/ 10.3390/waste1010012

Academic Editors: Vassilis Athanasiadis and Dimitris P. Makris

Received: 23 November 2022 Revised: 2 January 2023 Accepted: 4 January 2023 Published: 10 January 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Oyster shells are subjected to physical and chemical treatments before they can be used as high value-added resources [4]. However, their consumption as a calcium substitute is minimal, because only a meagre percentage is applied during the dark fermentation of cement waste [14,15]. Previous studies indicate that the oyster shell content in concrete cannot exceed 15% [4]; the suggested concentration in dark fermentation is less than 8% [15]. Thus, the limited application of oyster shells in the construction industry cannot fundamentally help in the recycling of the neglected oyster shells (NOSs). Oyster shells have also been used in water treatment, as they are considered as absorbents that help remove phosphorus and nitrogen compounds [11,15,16]. However, their absorption capacity is minimal at ~228.15 mg-P/kg-oyster shell, and the recovered absorptive materials are generated as secondary waste [17].

There is a massive demand for calcium compounds such as CaCO3 and Ca(OH)2 as they are commonly used for acid gas treatment; moreover, several studies have been conducted on SOx removal using CaCO3 [18,19]. This process involves two stages where CaCO3 reacts with SO4 to produce gypsum (CaSO4·2H2O) as the final product, as shown in Equation (1). Oyster shells can be used as a substitute for limestone, which would effectively address the over-exploitation of limestone in the environment.

$$\text{H}\_2\text{SO}\_4 + \text{CaCO}\_3 + 0.5\text{O}\_2 + \text{H}\_2\text{O} \leftrightarrow \text{CO}\_2 + \text{CaSO}\_4 + 2\text{H}\_2\text{O} \tag{1}$$

Interestingly, Lim et al. [18] found that the efficiency of flue gas desulfurization (FGD) increased with the addition of NaCl. SO2 removal increased up to 52% under Ca/S = 2 with the addition of 3.2% of NaCl, whereas it was relatively low at 32% in the absence of NaCl [20]. The sulfidation of CaO with alkali NaCl was much higher than that of CaO without alkali NaCl, as shown in Equation (2) [21]:

$$\rm H\_2SO\_4 + CaCO\_3 + 2NaCl \leftrightarrow Na\_2SO\_4 + CaCl\_2 + H\_2O + CO\_2 \tag{2}$$

However, when the sulfidation temperature was over 750 ◦C with the inclusion of sodium, the higher sodium concentration led to a greater spreading resistance ratio for sulfidation. As NaCl-CaSO4 partially melts under high temperature, the liquid phase present locally blocks the pores. In other words, a higher sodium concentration increases gas spreading resistance in the pores [22]. Furthermore, the gypsum (CaSO4·2H2O) recovered after FGD is high in sodium, which limits its conversion into viable products. Additionally, previous studies have reported that NaCl impurities affect the degree of limestone calcination [23–26]. Therefore, the use of oyster shells as a viable resource requires the removal of NaCl.

To overcome the existing limitations, in the present study, we investigated washing processes of oyster shells to convert them into useful resources.

#### **2. Materials and Methods**

#### *2.1. Oyster Shell Sample Collection*

Oyster shells were collected from an open-air yard in Tongyeong, South Korea, where shells are stored outdoors for over six months to generate shell fertilizer; this facilitates the natural washing out of NaCl from the shells. The shell separates the luggage rope using a cylindrical centrifuge, and in this process, some of the shell is crushed. The speed of the centrifuge is about 250 rpm, and the size of the crushed shell is about 5 cm or less. The NaCl content in these shells is expected to be lower than in those generated freshly at the shucking place.

#### *2.2. Oyster Shell Samples Subjected to Washing Process*

#### 2.2.1. Batch Washing

To identify the characteristics of NaCl removal from the oyster shells, an elution experiment was conducted using batch as well as sequential reactions. The batch experiment was conducted with 500 g of shells in a container by washing them with tap water, varying

the ratio of the volume of tap water from one to six times the weight of the shells; the different ratios of 1:1, 1:3, 1:4, 1:5, and 1:6 were denoted as B-OWR1 to B-OWR6 (B-OWR, batch-oyster shell to tap water ratio). Each oyster shell sample was agitated with the corresponding volume of tap water at 100 rpm at room temperature. Most of the experiments were conducted at 100 rpm, but the effect of rpm was investigated by changing the rpm to 200.

#### 2.2.2. Continuous Washing

In the continuous washing process, oyster shells were washed once, and the tap water used for elution of NaCl from the shells was removed, followed by the addition of new tap water. The washing and elution steps were repeated six times per sample. The samples were agitated for 1 min; the reacted solution was recovered after solid–liquid separation by filtration. The ratios of tap water used for the washing steps were maintained the same as those used for batch washing; the different ratios were denoted as C-OWR1 to C-OWR6 (C-OWR, continuous-oyster shell to tap water ratio). The number of elution processes was determined based on the concentration of NaCl obtained with each wash.

#### *2.3. Estimation of NaCl Leachates*

Considering that the oyster shells were not washed with tap water during shucking, the concentrations of leached dissolved ions from the shell could be associated with those of seawater components. The NaCl content of the tap water eluted after batch or continuous washing was estimated indirectly using electrical conductivity (EC) (Pro 30, YSI, Yellow Springs, Ohio, USA). From the measured EC values, the NaCl content in the shells was calculated according to Equation (3) [27]:

Total dissolved solids (mg/L) = k × Electrical conductivity (liquid − tap water) (μS/cm) <sup>×</sup> SSR (3)

where k is conversion of NaCl from the observed TDS value, that is 0.064, and SSR is the solution to solid ratio.

#### *2.4. Analysis of Oyster Shell Particle Size and Chemical Composition*

Shells recovered after washing were classified based on their particle size and chemical composition. The particle-size of the shells that ranged from 0.2 mm to 5 mm were classified using standard sieves. Chemical compositions of the shells before and after washing were identified using X-ray fluorescence spectroscopy (XRF-1800, Shimadzu, Kyoto, Japan), and an X-ray diffractometer (XRD-6100, Shimadzu, Kyoto, Japan) was used to analyze the mineral phases.

#### **3. Results and Discussion**

#### *3.1. Physical Properties of Neglected Oyster Shells*

This study evaluated the basic characteristics of NOSs near Tongyeong, South Korea. The NaCl content of NOSs was measured considering the number of repeated experiments, which were repeated more than six times at B-OWR6 for 24 h. As shown in Figure 1, the NaCl content of shells varied depending on the duration. From the result values, the average of the measured values repeated six times was used and the moisture content of sampled shells was measured as 1.22% ± 0.03 on average.

**Figure 1.** NaCl content of neglected oyster shells in B-OWR6 for 24 h.

Chemical analysis of the oyster shell samples using XRF indicated CaO, SiO2, Na2O, SO3, and MgO as the major constituents, and the average of the measured values repeated six times was used. As shown in Table 1, the presence of Na, Mg, and Mg in the shells suggests that major elements of the seawater could be captured when oysters grow in their habitat. Meanwhile, Al, Fe, and Si were believed to have originated from marine sediments [28,29].


**Table 1.** Chemical composition of neglected oyster shells (NOSs).

However, Cl was not detected in the XRF analysis. Because NaCl was regarded as the major salt, the XRF results suggest that NaCl was possibly washed out owing to its high water solubility (up to 36%). In fact, oyster shells in this study were obtained from a fertilizer production facility, where shells were left outdoors for more than six months, indicating that the high level of Cl in the shells had been washed away naturally, and hence, it was not detected.

#### *3.2. Removal of NaCl Using Washing Methods*

Since washing efficiencies could be attributed to the physical properties of the oyster shells, the washing method was tested based on the volume of water used, washing duration, and mixing speeds.

#### 3.2.1. Batch Washing

As previously mentioned, NaCl is highly soluble in water; therefore, its leaching would depend on the volume of water used in the washing step. Figure 2 illustrates the

leaching properties of NaCl from the shells with elapsed time depending on the volume of water, and the average of the measured values repeated three times was used.

**Figure 2.** Neglected-oyster shell to water ratio (OWR) depending on elution time. (**a**) NaCl elution concentration and (**b**) NaCl elution rate. B-OWR, batch-oyster shell to water ratio.

With respect to the volume of water, the absolute value of eluted NaCl was relatively higher when the water to shell ratio was higher (Figure 2a). However, it was observed that the elution volume of NaCl increased among samples with a smaller ratio of water used for washing (OWR < 4.0). In addition, the leaching of NaCl increased with an increase in elution time in general, except for the OWR of 5. However, no substantial effect was observed when a large ratio of water (OWRs of 5 and 6) was used. We observed a relatively higher absolute elution value when shells were washed for a relatively longer time at OWRs > 5 than that when washed at lower OWRs (Figure 2b). NaCl was eluted into the water during the initial reaction, which indicates that the concentration of eluted NaCl depends on the volume of water used for washing. Hence, it could be inferred that continuous washing for shorter periods was more effective as it enhances NaCl removal from the shells. The average NaCl content in the shells was 1.22 wt.%, and the removal efficiencies are shown in Table 2. The average of the measured values repeated three times was used. The batch system sufficiently removed NaCl residues from the shells under a B-OWR of ≥5. However, some differences persisted depending on the volume of washing water used.

**Table 2.** Maximum NaCl elution volume depending on water.


Figure 3 shows leaching of NaCl from the shells based on the mixing speed and the average of the measured values repeated three times was used. Leaching patterns of NaCl were very similar regardless of mixing speeds of 100 and 200 rpm. However, the amount of NaCl leached from the shell at 200 rpm was approximately 20% higher than that at 100 rpm. Leaching concentrations of NaCl at 200 rpm increased from the start of the reaction and eventually reached up to 80%. This indicates that over 90% of NaCl removal efficiency relies on the volume of water, even though leaching properties primarily rely on the mixing speed.

**Figure 3.** NaCl elution concentration (**left arrow**) with elapsed time at B-OWR3 with different mixing speeds. The bar chart is difference between values 100 rpm and 200 rpm (**right arrow**) with elapsed time at B-OWR3.

These results indicate that the volume of water used for washing is key for effective removal of NaCl from the surface of oyster shells. The removal efficiency of NaCl from the shells reached over 90% when the OWR was >5.0 (Table 2).

#### 3.2.2. Continuous Washing

Figure 4 shows the washing efficiencies for each washing step by continuous washing of oyster shell samples and the average of the measured values repeated three times was used. Leaching of NaCl from the shell was proportionate to the number of washing steps. The amount of leached NaCl at the first step was over 63–96% depending on the OWR. As shown in Figure 2 and Table 2, over 90% leaching was observed when the OWR was >5. Considering that the NaCl removal efficiency reached 90% by sequentially washing the shells twice even at a first OWR, a continuous washing method would help reduce the volume of water used for washing.

**Figure 4.** NaCl elution concentration via first and second batch system with various OWRs.

In the present study, we comparatively performed shell washing under different OWRs to reduce the reaction time. OWRs were chosen at two, three, and four times water volumes. The washing steps were repeated 10 times sequentially, as shown in Figure 5a, which also

shows the unit mass of leached NaCl from the shells at every single step. As confirmed by batch experiments (Figure 2), a high concentration of eluted NaCl was observed during the early stages of washing steps (up to three times) and reduced gradually reaching nearly zero after six washes. The leaching patterns observed were similar regardless of OWRs; however, the total amount of leached NaCl differed largely under varied OWRs. The average of measured values repeated three times was used, as shown in Figure 5b. C-OWR4 had the highest value of 11.1 g/kg-OS among the three samples, which is 2.2-fold higher than that of C-OWR2, implying that volume of water is the key factor that controls washing properties.

**Figure 5.** Amount of NaCl elution with continuous washing steps at different OWRs. (**a**) Changes in NaCl concentration as a function of washing steps. (**b**) Total NaCl elution concentration after 10 times of sequential washing.

We observed a substantial difference in NaCl leaching properties between the batch and continuous experiments. Assuming that the percentage of maximum eluted NaCl could be approximately 1.2%, leaching rates of NaCl at B-OWR5, B-OWR6, and C-OWR4 are repeated three time and indicated in Table 3. The maximum values (1.2%) of C-OWR4 were reached at 10 min, whereas batch experiments (B-OWR5 and B-OWR6) were 15 times more time consuming and utilized larger volumes of water to achieve the maximum elution value. As a result, the leaching rates of NaCl from the shells seem to be more affected by the volume of water than the duration of washing periods.


**Table 3.** Comparison of the optimum conditions between the batch and continuous systems.

#### *3.3. Particle Size of Washed Shells and Changes in Mineral Composition*

The particle size distributions are depicted in a parabolic shape, accounting for 70% of coarse grains over 5.0 mm, and the average of measured values repeated three times was used in Figure 6a. The particle size for raw shells indicated that particles ≥1 mm accounted for 78% and those ≥5 mm in the coarse-grains accounted for 70%.

**Figure 6.** Properties of particle size of washed neglected oyster shells (NOSs). (**a**) Particle size distribution. (**b**) Results of X-ray diffractometer (XRD) analysis of washed NOSs with different particle sizes. (**c**) CaO contents of washed NOSs with different particle sizes. (**d**) Al2O3, SiO2, and Fe2O3 contents of washed NOS with different particle sizes.

The NOSs were assessed using an XRD, and the patterns are represented in Figure 6b. Mineral composition of separated shells analyzed by XRD analysis confirmed that the main mineral was CaCO3 regardless of the particle size; some quartz was observed in the fine-grained particles [30]. These results were averaged three times and agreed with those of XRF; the main element found in particle sizes of over 0.2 mm was Ca, which accounted for >98% (Figure 6c). The content of CaO increased as the particles became coarser, and Si, Fe, and Al were relatively more enriched in the finer particles. Generally, these elements are the main components (Al2O3, SiO2, and Fe2O3) of the soil, suggesting that marine sediments could be recovered together with oyster shells from oyster aquacultural points. In Figure 6d, these results were averaged three times, the composition of Si, Fe, and Al were relatively more concentrated in the finer particles, and the content of these impurities clearly increased in particles ≤2 mm.

#### *3.4. Leaching Experiment*

Total component analysis results by elution test were compared for the washed and unwashed shells (Table 4). Although some differences were observed depending on the particle size, Na and Cl contents in washed shells clearly decreased by 77–93%. With the exception of Zn, heavy metals were not detected and decreased with increasing particle size. Some Na content remained after washing, which is likely because some seawater components are captured when oysters grow. Nonetheless, given that impurities in shells were relatively concentrated in particles, it is important to apply washing conditions considering the particle size to improve washing efficiency. Therefore, the sorting of particle size helps to remove impurities and NaCl simultaneously, as most of the impurities are confined to the finer particles. These results indicate that it is important to remove the impurities along with salts to use shells as Ca resources and selectively utilize fine-grained shells of ≥1.0 mm.


**Table 4.** NOSs and washed NOS components under different particle size by elution test.

#### **4. Conclusions**

The results of our study indicate the elution efficiency of NaCl with varied OWRs under the batch and continuous systems and different washing speeds and particle sizes. To completely elute out NaCl in one wash, at least five times as much water and 24 h mixing is required under the batch system. The NaCl elution is effective under a higher OWR where elution time can be minimized. The optimum NaCl elution efficiency of NOSs was predicted at an OWR of over 5, with an agitation time of 10 min and under a continuous washing system. Furthermore, because the NaCl proportion mainly constitutes fine particles, selective washing of particles ≥ 1.0 mm would increase washing efficiency. Larger particles of NOSs may benefit and be enhanced by the elution efficiency of NaCl. The washing speed is proportional to washing efficiency; otherwise, the elution efficiency according to the washing speed has a limitation in that the elution time would be longer in a batch system than in a continuous system. In conclusion, the washing process is essential for use neglected-oyster shells as a resource and materialize, and the use of the neglected oyster shells is a very important consideration for solve environmental problems.

**Author Contributions:** Conceptualization, S.K. and J.E.P.; formal analysis, data curation, S.E.L. and J.E.P.; writing—original draft preparation, writing—review and editing, J.E.P.; supervision, project administration, S.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by Environment Industry & Technology Institute (KEITI) through the Project to develop eco-friendly new materials and processing technology derived from wildlife Project, funded by Korea Ministry of Environment (MOE) (2021003280003).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

MDPI St. Alban-Anlage 66 4052 Basel Switzerland www.mdpi.com

*Waste* Editorial Office E-mail: waste@mdpi.com www.mdpi.com/journal/waste

Disclaimer/Publisher's Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Academic Open Access Publishing

mdpi.com ISBN 978-3-0365-9538-2