Hydrothermal treatment for inactivating some hygienic microbial indicators from food waste–amended animal feed

Abstract

To achieve the hygienic safety of food waste used as animal feed, a hydrothermal treatment process of 60–110 °C for 10–60 min was applied on the separated food waste from a university canteen. Based on the microbial analysis of raw waste, the inactivation of hygienic indicators of Staphylococcus aureus (SA), total coliform (TC), total aerobic plate counts (TPC), and molds and yeast (MY) were analyzed during the hydrothermal process. Results showed that indicators' concentrations were substantially reduced after hydrothermal treatment, with a greater reduction observed when the waste was treated with a higher temperature and pressure and a longer ramping time. The 110 °C hydrothermal treatment for 60 min was sufficient to disinfect food waste as animal feed from the viewpoint of hygienic safety. Results obtained so far indicate that hydrothermal treatment can significantly decrease microbial indicators' concentrations but does not lead to complete sterilization, because MY survived even after 60 min treatment at 110 °C. The information from the present study will contribute to the microbial risk control of food waste–amended animal feed, to cope with legislation on food or feed safety.

Implications:

Reduction of microbial indicators at ramping time and holding time during the hydrothermal process showed that hydrothermal treatment is an effective method to achieve hygienic feed from food waste to a certain extent, but the conditions researched in this study were not enough for the complete sterilization of food waste, because of the different heat resistance of bacteria and spores.

Introduction

In China, over 158 million tons of municipal solid waste was generated in 2010 (National Bureau of Statistics of China [NBSC], 2011). Of this, the food waste content is approximately up to 60% (Luo, 2010) on a wet-weight basis. A mass of food waste is generated from commercial service sites (such as hotel, restaurant, etc.) and educational, military, and health care institutions, which includes uneaten food and food preparation leftovers. It can be categorized as originating from four primary sources: raw material, food processing, post processing, and post consumer (Westendorf, 2000).

The large and increasing amount of food waste generated each year has raised concerns about its treatment and disposal. Among which, recycling of food waste for compost (Adhikari et al., 2008), energy (Zhang et al., 2007), or animal feed (Elferink et al., 2008; Westendorf, 2000) according to the quality of raw waste has expanded rapidly in recent years and this expansion is expected to continue, although at a slower rate. One important obstructive factor during the expansion is the complaints about disturbing odors and production safety from the public. These have been attributed to the activity of microbes, especially some pathogens. Therefore, it is necessary to develop and implement environmentally efficient methods for food waste management and treatment.

Waste disposal options, such as landfill or incineration, for food waste are becoming more expensive and scarce, due to some typical characters, such as high moisture content, high edible oil content, etc. On the other hand, most food waste is rich in nutritional ingredient and generally free from contamination before disposal (García et al., 2005; Westendorf, 2000), which is included in the category of regulated recyclable resources, especially for the garbage generated from restaurant (European Union [EU], 2009; Schapper and Chan, 2010; Tsai et al., 2007). Therefore, food waste has been reused in compost or livestock feed as a sort of recyclable resource. However, without a proper management, it is likely to be polluted through contact with disposal, transfer, or transport containers. Thus, almost 20 cities in China have made management regulations on the food waste generation, classification, collection, transportation, and processing to limit post-disposal contamination.

The concern of biological and chemical contaminants contained in waste has been focused on especially certain epidemiological situations (EU, 2009; García et al., 2005; Westendorf, 2000), which may limit its reutilization. In recent years, many countries such as Japan and China are facing with problems associated with some epidemic infections such as foot-and-mouth disease (Ma, 2010), poliomyelitis (Centers for Disease Control and Prevention [CDC], 2011), bovine spongiform encephalopathy, avian influenza (bird flu, H5N1) (World Health Organization [WHO], 2011), etc., which are caused by some typical pathogens such as picornavirus, enterovirus, prion protein, and avian virus. These disease-causing pathogens could potentially infect animal through recycled products (feed) ingestion and then be likely transferred to human consumers. This is of paramount concern in food waste reuse as animal feed. In order to control the biosecurity associated with microorganisms, some treatments are usually available, including boiling, chemical additives, ensiling, composting, or heat methods (EU, 2009; Kelley and Walker, 1999; Westendorf, 2000).

As a way to disinfect microbial, the use of heat is very common. Many technologies applied in food waste recycling in China are temperature dependent, and pathogen inactivation usually needs to be enhanced at high temperatures by different mechanisms. Among these, hydrothermal treatment involves applying heat under pressure to achieve reaction in an aqueous medium. It has been attracting worldwide attention because of the fascinating characteristics of water as reaction medium at elevated temperatures and pressures, for example, guaranteed processing in hygienic conditions (Rahman et al., 2004; Ren, 2006) and improved nutritional value of organism (He et al., 2008; Ren, 2006). Because of the distinctive characteristics described above, hydrothermal treatment is an effective method for the treatment of organic wastes, including food waste. Therefore, the hydrothermal processes become viable practices in China, such as the Suzhou Jiejin Food Waste Recycling Project (Ren, 2006).

Generally, hydrothermal treatment is used to sterilize pathogens under the environment of high relative humidity, and Rahman et al. reported that the flashing of moisture in a hydrothermal process tends to damage the cell further with increase of temperature in tuna mince (Rahman et al., 2004). But the researches on inactivation of food waste are few. Moreover, the characteristics of food waste, including the high moisture content of 74–90%, 3% oil content, and 4.59% salt content (based on dry weight with 79.01% moisture content), etc. (Ren, 2006), have an unavoidable influence on the microbial activity and inactivation (Prescott, 2001).

Therefore, our objective is to evaluate the effect of hydrothermal treatment on the reduction or elimination of selected microorganisms from food waste, improving its quality to be used in compost or livestock feed as a recyclable resource. Based on the analysis of total coliform (TC), total aerobic plate counts (TPC), molds and yeast (MY), staphylococcus aureus (SA), and Salmonella in raw food waste using standard culturing techniques, this study investigated the inactivation of indigenous indicator microorganisms such as TC, TPC, and MY during hydrothermal treatment. Simultaneously, the decimal reduction time (D-value) and the thermal resistance constant (Z-value) were determined for the inactivation test. Based on this analysis, this study can provide an initial insight into the ability of hydrothermal treatment processes to reduce the indigenous microbial risk associated with food waste directly.

Materials and Methods

Study description

The thermal inactivation of microbial indicators in food waste was carried out at 60, 80, and 110 °C. Pretreated food waste in the amount of 10–15 kg (wet weight) was poured into the experimental apparatus and then electrically heated. Based on the principle of heat transfer, the steam was generated from the waste and it took some time to reach the predetermined temperature. The thermal equilibration time was recorded when the target temperature was attained. A set of samples were collected from the apparatus immediately after the target temperature was attained, and more samples were collected after 10-min holding time intervals at the isothermal temperatures of 60 ± 2, 80 ± 2, or 110 ± 2 °C.

The microbial indicators of hygienic safety studied here originated from human food and animal feed. The disinfecting effectiveness was assessed by comparing the microbial counts of TPC, TC, SA, and MY in raw food waste before and after heat treatment. The microbiological analysis of the food waste always showed the absence of some pathogens such as Salmonella, enterobacteria (Esherichia coli), and molds and yeast (Viana and Schulz, 2003), but food waste is prone to contamination and is suitable to foster bacterial growth due to its high nutrient content. This suggests that pathogenic indicators could survive in the food waste even after the heat treatment and would be ingested by livestock to cause infectious diseases. Therefore, pre- and post-treatment waste samples were inoculated with standard cultures to test the efficiency of the hydrothermal treatment process to sterilize food waste using SA, TC, MY, and TPC as indicator organisms.

Substrate preparation

Raw food waste was collected from student cafeterias at Tsinghua University. Since the food waste was collected from original sources containing considerable impurities, mainly paper, plastics, and wood, manual sorting was used to prepare the food waste for treatment. The raw waste was separated to remove the coarse contaminants and then flowed into a mincer for size reduction and homogenization. The food waste used in this study was mainly composed of postconsumer plate scrapings containing uneaten food, a few waste paper fragments such as napkins and cups that were not successfully separated, and preconsumer unserved prepared foods. Its physicochemical properties are shown in Table 1. Of these, the organic matters, crude protein, ether extract, crude fiber, salt content, and floatable oil content were measured on the dry basis of the restaurant garbage.

Table 1. Physicochemistry properties of food waste samples (wt. %)

	Physicochemistry Properties ^†
	pH	Moisture	Organic Matters*	Crude Protein*	Ether Extract*	Crude Fiber*	Salt Content*	Floatable Oil *
Mean	6.67	77.18	91.07	28.64	31.45	3.05	2.68	10.2 × 10^−6
SD	0.33	4.22	4.87	3.37	2.13	1.89	0.59	2.23 × 10^−6
Notes: ^†Six samples were analyzed in triplicate. *Data measured on dry-weight basis of food waste.

For each sampling, approximately 0.5 kg (wet weight) of food waste was stored in 500-mL presterilized glass jars and plastic bags and was then sealed and delivered for microbial analyses as soon as possible. The samples were kept at 4 °C during transit until analysis. Samples of raw food waste and pretreatment waste were also collected aseptically and transferred into sterile glass jars prior to delivery for microbial analyses. One to three pre- and post-treatment samples were collected from each heat treatment batch, and triplicate samples were each analyzed twice to measure the sampling and analytical errors.

Experimental apparatus

Figure 1 presents a schematic illustration of the experimental apparatus used for hydrothermal treatment that mainly consisted of an electric motor, a steel heat reactor and a fuzzy PID (proportional-integral-derivative) controller. The reactor unit comprised the following parts: an electrically heated layer, a scraped-surface agitator, and a reaction chamber (effective volume 20 L, length 459 mm) with a switchable sealing valve to control the pressure in the vessel. The reactor could abide rough use and great heat as 220 °C, corresponding to a pressure of 2.0 MPa. The temperature and blender rotational speed were controlled by the fuzzy PID controller. The samples were collected from the bottom port.

Figure 1. Schematic diagram of the inactivation apparatus used for the hydrothermal treatment process.

The samples were disinfected in the reaction chamber by the heat transferred from the electrical heating layer. For hydrothermal treatment, the moisture content of pretreated waste should be over 80%. As the moisture content in food waste in China ranges from 75% to 95%, water should be added to or removed from the raw food waste by stirring or centrifugation before adding the waste to the apparatus. In this study, the food waste was adjusted to a moisture content of 85% before hydrothermal treatment.

Microbiological analysis

The microorganisms and analysis methods used in the present study were selected according to existing standards and recommendations of food and feed hygiene. They include total coliform (TC), total aerobic plate counts (TPC), molds and yeast (MY), as well as staphylococcus aureus (SA) and Salmonella (Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China [AQSIQ], 2001a, 2001b; Ministry of Health of the People's Republic of China [MOH], 2010). Of these indicators, TC, TPC, MY, and SA were counted using three-tube most probable number (MPN) method for agar, plate count agar, rose bengal medium agar, and Baird-Parker agar, respectively. Moreover, plasma coagulase test was done for SA identification (MOH, 2010). The identification of Salmonella was carried out by the Comprehensive Test Center of the Chinese Academy of Inspection and Quarantine with the approval of the China National Accreditation Service for Conformity Assessment based on the standard method (MOH, 2010). All agar media were purchased from Beijing Aoboxing Bio-tech Corporation (Beijing, China).

Statistical analysis

Results were expressed as the mean values and standard deviation of raw data. Data were expressed as log₁₀ colony-forming units (cfu)/g and analyzed using Excel (Microsoft Office 2007) and OriginPro7.5 (OriginLab Corporation, Northampton, MA, USA). Significant differences were determined by the Fisher's least significant difference (LSD) test at the P < 0.05 level.

Results and Discussion

Microbial concentrations of raw food waste

Table 2 lists the selected microorganisms that were inoculated and analyzed in the raw food waste, with the means and standard deviations of the counts. TPC and MY were detected in all 16 samples with mean counts (log₁₀ cfu/g) of 7.09 and 4.88, respectively, and maximal levels of 8.58 and 5.72, respectively. The concentration of TC in raw food waste ranged from below the limit of detection to 6.66 log₁₀ MPN/g wet weight (mean = 5.63), whereas the detectable rate was as high as 87.5%. SA was observed with a relatively high positive rate of 81.25%, and the frequent presence indicated poor hygiene in the raw food waste (Sancho et al., 2004). However, pathogenic Salmonella bacteria, which would form a healthy risk, could not be detected in the 16 original substrates, as reported by Viana and Schulz (Viana and Schulz, 2003). It indicated that the food waste researched in this study was free from contamination in the production source.

Table 2. Concentrations of indicator microorganisms in the raw food waste (based on wet weight)

	Microorganisms
	TPC (log10 cfu/g)	TC (log10 MPN/g)	MY (log10 cfu/g)	SA	Salmonella
Mean ± SD	7.09 ± 0.91	5.63 ± 0.91	4.88 ± 0.81	81.25% PR ^d	ND ^e
N ^a	16	14	16	13	0
MDL ^b	0	2	0	3	16
Total ^c	16	16	16	16	16
Notes: ^aNumber of samples included in the statistical analysis (mean and SD determination). ^bNumber of samples below minimum detection limits (10 cfu/g wet weight or 3 MPN/g). ^cTotal number of samples analyzed. ^dPositive rate of staphylococcus aureus. ^eNot detected above background.

Statistical analysis indicated that the standard deviation values of microbial concentrations were relatively high because of the wide variability among different samples (e.g., from 5.71 to 8.58 log₁₀ cfu/g wet weight TPC). Standard deviations of the replicates and duplicates of individual samples were typically at least 1 log₁₀ lower than the mean values (e.g., 1.45 × 10⁷ ± 1.14 × 10⁶ cfu/g wet weight TPC for one sample; N = 6). Hence, high variability arises from the heterogeneity of the samples tested rather than inconsistent analysis techniques. This also applies to the following results.

Removal of microbial indicators in food waste by hydrothermal treatment

Removal of each microbial indicator associated with hygienic safety in food waste was investigated under the hydrothermal treatment condition at 60, 80, and 110 °C. To calculate logarithmic values, 1 was used when the indicator's concentration was below the limit of detection, and SA was not included in this analysis because of the qualitative nature of this detection. All samples were negative for SA after hydrothermal treatment for 30 min at 80 °C.

Figure 2 shows the concentration of TPC and MY in food waste during the experiment period. Both of TPC and MY tended to exhibited two stages, i.e., an initial stage with slowly decrease during the ramping time, especially in the low temperatures such as 60 and 80 °C. After the apparatus reached the target temperature, a rapid inactivation was found until the indicators became undetectable except MY following an extended holding time (Figure 2).

Figure 2. Survivor curves for indigenous microbial indicators in food waste were determined by various thermal (60, 80, and 110 °C) conditions. The dotted vertical line indicates the ramping time. The dash horizontal line means at least 3 decimal reductions of TPC, whereas the dash dot horizontal line shows the 3 decimal reductions of MY. (▪) TPC; (○) molds and yeast.

Reduction during ramping time

TPC and MY indigenous indicators differed in their resistance to hydrothermal treatment during the ramping time (Figure 2 and Table 3). The TC bacteria were susceptible to heat even at 60 °C, with the logarithmic reduction reaching 99% during the ramping time (data not shown).

Table 3. Typical experimental conditions and logarithmic reduction of indigenous microorganisms during ramping time of the hydrothermal processing

			Microorganisms
Process Temperature (°C)	Process Relative Pressure (MPa)	Ramping Time (min)	TPC (log10 cfu/g)	MY (log10 cfu/g)	SA
60 ± 2 *	0.06	12	0.02 ± 0.03 ^a	0.01 ± 0.01 ^a	100% PR* *
80 ± 2*	0.12	20	0.04 ± 0.03 ^a	0.03 ± 0.02 ^a	100% PR* *
110 ± 2^+	0.20	60	0.92 ± 0.14 ^b	0.31 ± 0.24 ^b	100% PR* *
Notes: ^a,bMean (± standard deviation), with different superscripts within a column indicating a significant difference (P < 0.05). *No significant difference between TPC and MY within a row (P > 0.05). ^+Significant difference between TPC and MY within a row (P < 0.05). **Positive rate of staphylococcus aureus.

For the other indicators tested, no significant reductions were observed during the ramping time for hydrothermal treatment at 60 and 80 °C under these experimental conditions (Table 3). However, when the temperature was increased to 110 °C, these bacteria exhibited significant log reductions (P < 0.05) during the ramping time (Table 3). TPC and MY were sensitive to thermal treatment at 110 °C, with the initial populations decreased by 0.92 and 0.31 log₁₀ cfu/g during the ramping time, respectively (Table 3). This shows that the temperature is extremely important for the removal of the indicators.

Meanwhile, the effect of the heat time and pressure should not be ignored. Approximately 12 min was required to raise the temperature of apparatus to 60 °C, whereas a longer time was required to reach higher temperatures (Table 3). The ramping time to reach 80 °C was 20 min, whereas it took 60 min to reach 110 °C. Regardless of this discrepancy in ramping time, greater inactivation was observed at higher pressure in combination with higher temperature (Table 3). When 110 °C was reached, the absolute pressure in the apparatus was almost 0.3 MPa, which is the lower-limit pressure for pressure sterilization according to Regulation (EC) No. 1069/2009 (EU, 2009). Therefore, greater reduction was detected at 110 °C compared with 60 and 80 °C, respectively.

Moreover, different reductions of microbial indicators in food waste highlight the importance of reporting the ramping time and corresponding log reduction during hydrothermal treatment. In this study, the ramping time was necessary to carry out hydrothermal treatment because of the thermal conductivity of apparatus and food waste, as well as the actual practice of food waste treatment in China (Ren, 2006). And the corresponding logarithmic reduction of indicators during the ramping time was unavoidable. Thus, the ramping time and corresponding log reduction are significant for the practical engineering application of hydrothermal treatment.

Inactivation during process holding time

A reduction in all of the indicators occurred during the process holding time by heat inactivation associated with a higher temperature and pressure, in the experimental range of 60, 80, and 110 °C (Figure 2).

Among the sanitary indicators tested, MY was the most heat resistant. When TC and SA bacteria were treated at 80 °C and 0.12 MPa for 30 min of holding time, no viable bacteria were detected in the enrichment cultures (date not shown), whereas viable TPC and MY were detected in enrichment cultures after 60 min of treatment (Figure 2a). Similarly, MY appeared to be more heat resistant than TPC. Although food waste was treated at every experimental temperature for 60 min, more viable MY was detected in the enrichment cultures than TPC. More than 0.5 log₁₀ cfu/g of MY survived after 60 min of treatment at 110 °C, whereas TPC treated under similar conditions were not detected by the enrichment techniques (Figure 2c). According to the heat tolerance exhibited by indigenous indicators, monitoring TPC and MY densities for indicator purposes must be undertaken to achieve the hygienic safety of food waste treatment.

Moreover, although MY population required severe hydrothermal treatment of 110 °C–0.2 MPa for 60 min to achieve at least 3 decimal reductions, 80 °C–0.12 MPa for 30 min holding time was sufficient to eradicate a similar population of TPC (Figure 2b and c). According to microbiological specifications for hygienic safety of animal feed (Table 4) and the maximal levels of microbial concentrations in raw food waste (Table 2), it is assumed that no detectable Salmonella, SA, and TC as well as at least 3 decimal reductions of MY and TPC are necessary for the hygienic safety of food waste used as feed. Based on this assumption, 110 °C for at least 60 min hydrothermal treatment of food waste assures enough microbiological reduction to be considered innocuous as animal feed.

Table 4. Microbiological specifications for hygienic safety of animal feed

	Microorganisms
	TPC ^#	MY ^#	Salmonella ^#	SA ^§	TC ^†
Microbiological specifications	2 × 10^6 cfu/g	2 × 10^4 cfu/g	No detection	10 cfu/g	No detection
Notes: References: ^# AQSIQ (2001a); ^§ MAPA (1988); ^† AQSIQ (2001b).

Therefore, as a method of heat sterilization, hydrothermal treatment is an effective means for disinfecting food waste based on the theory of autoclaving. However, higher temperature and longer time are needed compared with the usual standard autoclaving condition of 121 °C for 15 min (Todar, 2009), which are caused by the large volumes of food waste (almost 10–15 kg), lower inactivation efficiency, as well as the protective effect to microbes in food waste. The apparatus of hydrothermal treatment used in the study operated by using steam under pressure as the sterilizing agent, just like autoclave, but the steam came from the vaporization of water of food waste. It took some time to convert boiling water to steam, and the productive steam wasn't as plenteous as autoclave. Furthermore, the heterogenized food waste containing some organic compounds with certain viscosity would form a protective shield around the microbes (Prescott, 2001). These circumstances make the hydrothermal treatment have similar function to autoclaving, but higher temperature and longer time are required.

However, survival of some indicators, for example, molds, yeast, as well as some epidemic pathogens, in food waste may be of certain concern such as causing poor policy stability (Lin et al., 2011), due to the possibility of survival and subsequent germination of heat-resistant spores and transfer of infectious disease. Additionally, the food waste composition is not always stable, which has an unavoidable influence on the microbial inactivation. These issues should be further focused on. On the other hand, the elimination of pathogen indicators such as coliforms during the process makes this prospect unlikely due to increased survival of coliform compared with most of the other pathogens of concern in feed safety.

Inactivation kinetics of TPC and MY during hydrothermal treatment

To investigate the relationship between the indicators and the various temperatures of 60, 80, and 110 °C, inactivation tests were conducted without considering the temperature ramping time and its associated bacterial lethality (Table 3). Generally, the heat destruction of microorganisms was analyzed based on the estimation of the decimal reduction time, which states that the time required for 90% inactivation (1 log reduction) of a microorganism at temperature T (K) is defined as D (decimal reduction time). D was estimated from the linear slope of log₁₀ (survival) versus temporal regression data for the initial portion (Rahman et al., 2000), as shown in Figure 3. The values of D were 19.6, 12.5, and 8.0 min and 28.6, 18.9, and 15.4 min for TPC and MY at 60, 80, and 110 °C, respectively. These values were significantly different from those found by Rahman et al. in tuna mince for moist heat treatment (Rahman et al., 2004). This variation suggests that D-values may vary for similar treatment methods depending on the source of the raw material. Additionally, the D-vaule of TPC was much lower than that for dry heat treatment, which was reported as 8.33 hr at 60 °C by Rahman et al. (Rahman et al., 2000), demonstrating that hydrothermal treatment led to greater bacterial inactivation than dry-heating treatment. As expected, the values found in the present study decreased with increasing temperature, indicating that increasing the temperature increased the lethal effect.

Figure 3. Plot of logarithmic survival for indigenous bacteria in food waste subjected to various thermal conditions. (○) 60 °C, (▪) 80 °C, and (Δ) 110 °C.

The effect of temperature is determined by the relationship between D and T, calculated using eq 1:

(1)

where D ₁ and D ₂ are the values of D at T ₁ or T ₂, and Z is the change in temperature corresponding to a 1 log change of D.

Hence, the values of log D are plotted in Figure 4 against temperature to estimate the Z-value, which indicates the resistance to continuous temperature changes. The results showed that the Z-value of TPC (129) was smaller than that observed for MY (185). This means that TPC was more sensitive to continuous temperature increase, but MY exhibited greater tolerance than TPC at each temperature (Figure 2). The difference of inactivation between MY and TPC by this hydrothermal process indicated that inactivation mechanisms varied among different indicators (Desai and Varadaraj, 2010). Generally, vegetative cells are more sensitive to heat when water evaporates, whereas spores are not (Rahman et al., 2004). This supports the view that D- and Z-values should not be used to guarantee biological safety, as sporulating pathogens, if present, could not be completely destroyed during the treatment. Additional mechanistic studies are essential for a better understanding of the fate of bacterial spores under hydrothermal conditions.

Figure 4. Plot of log D versus temperature.

Conclusions

Due to the abundant nutrition and easily deteriorated items of food waste, heat treatments are required to ensure the biosecurity for possible use. Results of this study indicated that hydrothermal treatment substantially reduced concentrations of microbial indicators in food waste, but did not lead to complete sterilization. From the viewpoint of hygienic safety, 110 °C for at least 60 min was sufficient to disinfect food waste by hydrothermal treatment. Furthermore, the increased temperatures and longer times generally resulted in decreased concentrations of indicators during ramping time or holding time in posttreated samples. Anyhow, the hydrothermal treatment process substantially decreased microbial indicators' concentrations and therefore reduced the potential of infectious disease transmission by recycling food waste.

Acknowledgments

This work was supported by the China National Key Project of Scientific and Technical Supporting Programs of China (no. 2009BAC64B06), and the Public Science and Technology Research Funds Projects of Environmental Protection, Ministry of Environmental Protection of the People's Republic of China (no. 201109035).