Delaware and Rhode Island permitted courts to waive requirements to pay existing legal obligations. Colorado vacated underlying convictions as part of its expungement process, and voided remaining fees, fines, and restitution associated with vacated cases.Among the 39 states and Washington DC that decriminalized or legalized cannabis use, we found that 36 permitted some type of expungement of conviction records related to cannabis and four states prohibited the expungement of existing convictions. Within the 36 states that allowed expungement, 34 offered general expungement, 21 offered cannabis-specific expungement, and 11 offered general drug offense expungement . Among the 34 states with general expungement programs, 33 allowed expungements by petition, 9 expunged records automatically , and 7 allowed pardons as part of records expungement. Among the 21 states with cannabis specific expungement programs, 17 allowed expungements by petition, 9 expunged records automatically, 5 allowed pardons, and 7 provided multiple expungement programs. States with petition systems require applicants to navigate complex procedures that consume time and resources . Administrative burdens are broadly defined as the compliance, learning, and psychological costs endured by persons interacting with civic institutions that are correlated with reductions in uptake of government services . Petition systems increase administrative burdens, and as a result navigating them may require record holders to secure legal assistance,cannabis grow systems further increasing costs and limiting relief for people who are socioeconomically disadvantaged due to their convictions . Cannabis laws have imposed disproportionate criminal burdens on minorities,making accessible expungement important to improving racial equity outcomes as cannabis legalization expands .

Limited knowledge of expungement, the education needed to navigate petition systems, stress incurred during the process, the requirement to wait a specific period before seeking relief, and the payment of fees likely contributes to limited use of petition-based expungement . State policymakers may create, or retain, systems based on petitions that make expungement more difficult in the belief that this will encourage lawful behavior, and that it will prevent those who are likely to reoffend from having their records expunged . However, research suggests that expungement reduces recidivism , improves earnings, and increases employability by removing criminal records that limit the ability to obtain work, housing, or secure education funding . Expungements are also low-cost relative to job training programs and improve economic outcomes, as persons with criminal records earn less over time . Our findings suggest that potential beneficiaries may have been denied this economic and social relief, since the majority of states require record holders to petition for expungement. The majority of expungement programs also imposed administrative burdens in the form of waiting periods and fees. Administrative burdens can compound racial disparities associated with stigma, and even small administrative burdens can limit access to government aid and services . For example, waiting periods established by some states, in the expectation that these will ensure only “truly reformed” persons obtain record relief, can perpetuate inequalities as racial minorities are disproportionately criminalized by cannabis offenses . Among the states offering expungement, all but two had waiting periods, which varied in duration by offense levels and types. Waiting periods are administrative burdens because they create a psychological hurdle and sustain stress related to criminal record stigma and penalties . Shorter waiting periods improve record holders’ long-term earnings , while longer waiting periods can increase recidivism by hindering reentry into employment. Some states have reduced waiting periods to improve access to expungement . Waiting periods could arguably be eliminated for former cannabis offenders in states where use is legal and or decriminalized, given that criminalization has already been overturned .

We also found that most states required applicants to pay administrative fees and resolve outstanding judgments and legal financial obligations, pay restitution, or all of the above. Fees constitute another burden, either directly by increasing financial costs or indirectly by requiring applicants to expend time to learn how to navigate indigency waiver processes. We were unable to find prior research assessing the impacts of administrative fees on cannabis-related record holders or applicants for expungement. However, legal financial penalties disproportionately burden minorities and persons of lower socioeconomic status and are associated with increased poverty . Although multiple states permit courts to grant indigency waivers, judges do not always allow this assistance . Many record holders are unaware of indigency waivers, creating an invisible barrier to expungement . The inability to pay financial penalties may also inflict collateral consequences, such as driver’s license revocation or having debt reported to credit agencies . States could reduce these harms by eliminating administrative fees for expungement and postponing or eliminating payment of outstanding legal financial obligations. Previous research has suggested that states explore automated expungement systems to reduce administrative burdens and increase rates of expungement . Reducing administrative burdens has historically increased records expungement . Nonetheless, further research is needed on automatic expungement programs. For example, California’s historic petition-based expungement program for cannabis offenses resulted in 5% to 7% of eligible candidates applying . The state implemented automated expungement for cannabis offenses in 2018 to provide greater relief , but progress on automatically clearing the 220,000 cannabis conviction records estimated to qualify for clearance has been uneven . As of July 2021, 34,000 records remained unprocessed, spurring the introduction of new legislation to provide further automatic relief by January 1, 2023 .Our study has limitations. It focused on expungement for cannabis offenses in states that decriminalized or legalized cannabis use and sales by September 2022 and cannot be generalized to other offenses. People with cannabis offenses are more likely to secure records expungement than people with non-cannabis offenses, given the growing social permissiveness around cannabis and the expansion of legalization.

States also have expungement practices that this study did not capture, such as filing and dissemination requirements for expungement petitions and orders, mandatory hearings, imposing a burden of proof on applicants, or program qualifiers limiting expungement to either first-time offenders or persons that are diverted to and complete probation or drug treatment. In addition, we did not explore regulations that prevent legal aid organizations from filing petitions or other factors that could increase or reduce expungement access, such as prohibiting relief for people with unrelated offenses, imposing caps on total offenses, or on certain types of offenses. These requirements are administrative burdens that further complicate the process of records expungement. Although we cataloged expungement provisions pertaining to violations, misdemeanors, and felonies, we did not survey the levels and types of cannabis offenses in each state or how those may have changed over time . Instead, our paper surveyed all potentially applicable statutes and restrictions that could apply to a cannabis offense. Finally, cannabis is a dynamic policy area and legalization laws as well as expungement statutes have changed rapidly since 2012. Nonetheless,ebb and flow tables this research suggests how current expungement programs may affect the ability of cannabis-related record holders to secure relief in states that have attempted to reduce cannabis criminalization, and provides the groundwork for further exploration.Warehouse storage occupancies are currently reaching heights on the order of 24 to 30 m high stacks of storage commodity, which have not been considered by existing fire codes and engineering correlations. Over the last 50 years, fire protection engineers have relied on large-scale tests to classify commodities into one of seven classes that are representative of their fire performance under specific geometric configurations and ignition conditions. This classification process, which relies on expensive full scale testing, results in increased safety gaps as the industry creates new and untested materials that are stored in large quantities. Only a limited amount of fundamental science has been performed in this area, which is largely due to the range of complexities that occur in large-scale fire phenomena. Some correlations for large-scale flame heights of some commodities and fuels are present in the literature, but they are limited to specific fuels/configurations and some of the correlations require heat release rate values from full-scale tests. These correlations will be discussed further in the following sections of this paper. Currently, no tests that are known to the authors provide a complete set of fundamental, non-dimensional parameters that can be used in engineering calculations towards the safer design of large storage facilities. Efforts that result in the development of such test methods and classification methodology with a sound scientific basis may fulfill an urgent need to improve upon the current warehouse design methods. In most of these incidents, which were reviewed in Part I of this paper, the facilities were protected by automatic sprinkler systems that were installed in accordance with their respective current codes and standards. The negative impacts of these devastating fire incidents were felt by the occupants, firefighters, insurance interests, and local environments. From a business aspect, millions of dollars of materials or products are lost, and operations may be halted. Furthermore, insurance premiums are increased as a result of the fire, and the lost time can never be recovered. From a life-safety aspect, the lives of workers and responding firefighters are endangered, which can result in injuries or death.

The water runoff from firefighting operations and the resulting smoke plumes can also adversely affect the surrounding environment. The development of an approach to protect these facilities based upon the combustible materials that are stored, the layout of these materials, and the complex interaction with potential suppression systems is a critical step towards reducing the amount of devastating warehouse losses. As a first step towards improving the methods for commodity classification, two existing non-dimensional parameters are used to represent the physical phenomena present at the large-scale and model one part of large-scale flame spread. In Part I of this paper, a method was developed to experimentally quantify the burning rate of a material based upon the non-dimensional comparison of a materials chemical energy released during the combustion process with the energy required to vaporize the fuel, which was measured as a B-number. Commodities are classified to design sprinkler protection systems for most warehouse scenarios, and because such a sprinkler systems goal is to suppress or control a fire, the ranking of materials based upon the burning and spread rates of a potential fire is appropriate. Experiments were performed on a standard warehouse commodity, a Group A plastic, which is typically used to represent the worst-case commodity in large-scale tests. The commodity consisted of a single corrugated cardboard box that measured 53 x 53 x 51 cm and contained 125 crystallized polystyrene cups that were segregated by corrugated cardboard dividers. All of the faces except for the front face of the commodity were uniformly insulated, and the front face of the commodity was ignited at its base. The experimental observations of the Group A plastic commodity resulted in a qualitative description of the burning process over three distinct stages of burning. The first stage was characterized by upward flame spread over the front face of the corrugated cardboard, followed by a decreased burning rate as the cardboard smoldered and the polystyrene heated, and finally a sharp increase in the burning rate after ignition of the polystyrene. Despite the complex configuration, each stage resulted in distinct material involvement, which indicates the potential to model distinct material involvement from each stage using parameters derived from bench-scale testing. Fluctuations between the repeated tests also indicated the difficulty in obtaining repeatable measurements during these larger tests; therefore, small-scale test methods that can be repeated at a level of statistical accuracy may greatly improve the applicability of the results. Part II of this study continues the development of a non-dimensional approach to characterizing the burning behavior of materials. The bench scale tests that were performed in this study involved a small, flat sample of corrugated cardboard or polystyrene oriented vertically in which the burning was isolated to the front surface of the sample. The flow was considered to be laminar due to the observed behavior of the flow in the experiments. At the bench scale, the tests captured the effects of the commodity material properties on the flame spread process while separating the large-scale effects such as turbulence and radiation. Non-dimensional B-numbers were experimentally determined for the samples with greater accuracy than previous experiments. A flame spread model was then utilized to demonstrate the application of the experimentally measured B-numbers to predict in-rack flame heights in large-scale configurations.