Search for: All records

Various technologies and strategies have been proposed to decarbonize the chemical industry. Assessing the decarbonization, environmental, and economic implications of these technologies and strategies is critical to identifying pathways to a more sustainable industrial future. This study reviews recent advancements and integration of systems analysis models, including process analysis, material flow analysis, life cycle assessment, techno-economic analysis, and machine learning. These models are categorized based on analytical methods and application scales (i.e., micro-, meso-, and macroscale) for promising decarbonization technologies (e.g., carbon capture, storage, and utilization, biomass feedstock, and electrification) and circular economy strategies. Incorporating forward-looking, data-driven approaches into existing models allows for optimizing complex industrial systems and assessing future impacts. Although advances in industrial ecology–, economic-, and planetary boundary–based modeling support a more holistic systems-level assessment, more effects are needed to consider impacts on ecosystems. Effective applications of these advanced, integrated models require cross-disciplinary collaborations across chemical engineering, industrial ecology, and economics.

Expected final online publication date for the Annual Review of Chemical and Biomolecular Engineering , Volume 15 is June 2024. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

 

Replacing fossil fuels with biofuels offers a promising path to decarbonizing the transportation sector, a major source of greenhouse gas emissions (GHGs). Utilizing waste biomass such as forest residues is particularly appealing, as it avoids land-use change and associated GHG emissions. Current biofuel life cycle assessment (LCA) adopted by regulatory agencies considers forest residues as carbon-neutral feedstock and typically ignores soil organic carbon (SOC) changes from residue removal. Our study quantifies SOC change caused by removing forest residues in the Southern US and found that they can make a substantial contribution to the carbon footprint of biofuel derived from forest residues. Our results emphasize the need to include soil carbon assessment in future LCAs, biofuel policy, and forest management, even when waste biomass is used and no land-use change is involved. 

Cation exchange is a versatile post-synthetic method to explore a wide range of nanoparticle compositions, phases, and morphologies. Recently, several studies have expanded the scope of cation exchange to magic-size clusters (MSCs). Mechanistic studies indicated that MSC cation exchange undergoes a two-stage reaction pathway instead of the continuous diffusion-controlled mechanism found in nanoparticle cation exchange reactions. The cation exchange intermediate, however, has not been well-identified despite it being the key to understanding the reaction mechanism. Only indirect evidence, such as exciton peak shifts and powder x-ray diffraction, has been used to indicate the formation of the cation exchange intermediate. In this paper, we investigate the unusual nature of cation exchange in nanoclusters using our previously reported CdS MSC. High-resolution mass spectra reveal two cation exchanged reaction intermediates [Ag2Cd32S33(L) and AgCd33S33(L), L: oleic acid] as well as the fully exchanged Ag2S cluster. Crystal and electronic structure characterizations also confirm the two-stage reaction mechanism. Additionally, we investigate the Cu/CdS MSC cation exchange reaction and find a similar two-stage reaction mechanism. Our study shows that the formation of dilutely exchanged intermediate clusters can be generally found in the first stage of the MSC cation exchange reaction. By exchanging different cations, these intermediate clusters can access varying properties compared to their unexchanged counterparts. 

Afforestation and reforestation (AR) on marginal land are nature-based solutions to climate change. There is a gap in understanding the climate mitigation potential of protection and commercial AR with different combinations of forest plantation management and wood utilization pathways. Here, we fill the gap using a dynamic, multiscale life cycle assessment to estimate one-century greenhouse gas (GHG) mitigation delivered by (both traditional and innovative) commercial and protection AR with different planting density and thinning regimes on marginal land in the southeastern United States. We found that innovative commercial AR generally mitigates more GHGs across 100 y (3.73 to 4.15 Giga tonnes of CO 2 equivalent (Gt CO 2 e)) through cross-laminated timber (CLT) and biochar than protection AR (3.35 to 3.69 Gt CO 2 e) and commercial AR with traditional lumber production (3.17 to 3.51 Gt CO 2 e), especially in moderately cooler and dryer regions in this study with higher forest carbon yield, soil clay content, and CLT substitution. In a shorter timeframe (≤50 y), protection AR is likely to deliver higher GHG mitigation. On average, for the same wood product, low-density plantations without thinning and high-density plantations with thinning mitigate more life cycle GHGs and result in higher carbon stock than that of low-density with thinning plantations. Commercial AR increases the carbon stock of standing plantations, wood products, and biochar, but the increases have uneven spatial distributions. Georgia (0.38 Gt C), Alabama (0.28 Gt C), and North Carolina (0.13 Gt C) have the largest carbon stock increases that can be prioritized for innovative commercial AR projects on marginal land. 

Key Points Group 1 alkenones are reliable indicators of cold‐season temperatures Pre‐industrial cold‐season warmth between 1750 and 1850 CE in northeastern China Relatively warm cold season may be related to positive cold‐season Arctic Oscillation conditions 

Due to a growing interest in deep learning applications [5], compute-intensive and long-running (hours to days) training jobs have become a significant component of datacenter workloads. A large fraction of these jobs is often exploratory, with the goal of determining the best model structure (e.g., the number of layers and channels in a convolutional neural network), hyperparameters (e.g., the learning rate), and data augmentation strategies for the target application. Notably, training jobs are often terminated early if their learning metrics (e.g., training and validation accuracy) are not converging, with only a few completing successfully. For this motivating application, we consider the problem of scheduling a set of jobs that can be terminated at predetermined checkpoints with known probabilities estimated from historical data. We prove that, in order to minimize the time to complete the first K successful jobs on a single server, optimal scheduling does not require preemption (even when preemption overhead is negligible) and provide an optimal policy; advantages of this policy are quantified through simulation. Related Work. While job scheduling has been investigated extensively in many scenarios (see [6] and [2] for a survey of recent result), most policies require that the cost of waiting times of each job be known at scheduling time; in contrast, in our setting the scheduler does not know which job will be the K-th successful job, and sojourn times of subsequent jobs do not contribute to the target metric. For example, [4, 3] minimize makespan (i.e., the time to complete all jobs) for known execution times and waiting time costs; similarly, Gittins index [1] and SR rank [7] minimize expected sojourn time of all jobs, i.e., both successfully completed jobs and jobs terminated early. Unfortunately, scheduling policies not distinguishing between these two types of jobs may favor jobs where the next stage is short and leads to early termination with high probability, which is an undesirable outcome in our applications of interest.