Experimental and modeling study for hydrothermal liquefaction of wet biowaste

Aierzhati, Aersi

Permalink

https://hdl.handle.net/2142/108308 Copy

Description

Title

Experimental and modeling study for hydrothermal liquefaction of wet biowaste

Author(s)

Aierzhati, Aersi

Issue Date

2020-05-05

Director of Research (if dissertation) or Advisor (if thesis)

Zhang, Yuanhui

Doctoral Committee Chair(s)

Zhang, Yuanhui

Committee Member(s)

• Cai, Ximing
• Davidson, Paul
• Sharma, B.K.

Department of Study

Engineering Administration

Discipline

Agricultural & Biological Engr

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Hydrothermal Liquefaction
• Food Waste
• Techno-economic Analysis
• Pilot Scale Continuous Reactor
• Prediction Model

Abstract

The objective of this study is to demonstrate potential large-scale application of wet biowaste hydrothermal liquefaction (HTL) process. Experiments in lab scale and pilot scale are conducted to investigate the reaction mechanism and overcome technical challenges. Technoeconomic analyses are performed to understand the economic feasibility of the process. The lab scale study conducts HTL with elevated temperatures (280-380°C) and times (10-60 min) to convert categorized food residues into biocrude oil. The effects of different biochemical compositions of the feedstock on the biocrude oil production, energy balance, elemental distribution, and reaction mechanism are investigated. The mixture food waste yields 46.9 wt.% of biocrude oil at the optimal reaction condition. Eight distinct food waste feedstocks with different biochemical compositions (protein, carbohydrate, and lipid) results in significantly different biocrude oil yield. The influence of biochemical composition of feedstocks on the biocrude oil yield is ranked as lipid > protein > carbohydrate for the respective optimized HTL temperatures and times. A descriptive HTL process energy recovery and consumption ratio analysis show that feedstock with a higher lipid content has a higher energy recovery and lower energy consumption ratio (ECR). The food waste mixture generates an energy recovery of 72.2% and ECR of 0.23. This study and provides a complete description of elemental distributions (C, N, and P) based on feedstocks, optimized reaction conditions, and different phases of HTL products. These results are essential indicators for nutrient value recovery from food waste HTL products. GC-MS analysis of biocrude oil demonstrates that the most accumulated group of compounds are saturated fatty acids, monounsaturated Fatty Acids (MUFAs), polyunsaturated Fatty Acids (PUFAs), and fatty acid amides and esters. The first comprehensive network of reaction pathways for food waste HTL is developed base on these results. In addition, food waste HTL conversion prediction models are developed for biocrude oil yield and oil heating value. A second order polynomial model with multiple key variables of the process performs accurate prediction on food waste HTL biocrude oil yield. The model can accurately determine biocrude oil yield (R2 98.3%) of different food waste under different reaction conditions, as well as predict previously published data (R2 94.4%). For biocrude oil heating value prediction, different machine learning models are compared in R. The polynomial regression model is proven to be the best among these prediction models. This study also demonstrates a proof-of-concept in the production of renewable biofuel from food waste via HTL. Process model via ASPEN software and economic model with Microsoft Excel are built to analyze the technoeconomic feasibility of HTL and associated wastewater treatment process. Calculations on detailed capital cost and operating cost of these three major modules show that the HTL equipment cost accounts for about 32% of the total capital investment (TCI). The minimum selling point (MSP) of the base case (1000 dry kg per hour of food waste) is calculated to be $3.48/ Gallon of gasoline equivalent (GGE), which is higher than the long-term plan goal ($3.00/GGE) established by the Department of Energy. Sensitivity analysis on a range of essential manufacturing and financial parameters show that biocrude oil yield has the highest influence on the MSP. Reaction scale and feedstock price are another two big factors, followed by government policies, including tax incentives, green fuel subsidies, and overall discount rate. A 35-liter one-wet-ton-per-day pilot scale plug flow continuous HTL reactor for food waste is upgraded for demonstration purpose. The reactor operates at 300°C, and 1450 psi. The flow rate is controlled at 0.15 gal·min-1. Solid content of feedstock slurry is 13.3%. The pilot scale reactor results in 29.5 wt.% biocrude oil yield. GC-MS analysis shows the biocrude oil product from the pilot scale reactor has similar properties as lab scale HTL biocrude oil. However, the pilot scale biocrude oil molecular weight is higher according to TGA and MALDI analysis. Different business scenarios are compared between transporting the feedstock to the HTL or transporting the biocrude oil to refinery. A comparison of techno-economic analysis (TEA) shows that because of the lack of scale effect, the economics of the mobile HTL process is not as good as the onsite HTL process. However, if consider the transportation cost of feedstock and biocrude oil, at a distance more than 106 miles, the mobile HTL process is more economically feasible. In sum, this study showcases a technically cohesive and economically feasible approach to produce renewable biofuels from wet biowaste source for potential commercial applications.

Graduation Semester

2020-05

Type of Resource

Thesis

Permalink

http://hdl.handle.net/2142/108308

Copyright and License Information

Copyright 2020 Aersi Aierzhati

Owning Collections