monte carlo simulation tools crystal ball

#MONTE CARLO SIMULATION TOOLS CRYSTAL BALL SOFTWARE#
#MONTE CARLO SIMULATION TOOLS CRYSTAL BALL TRIAL#

Model for tobacco dry weight estimation as a function of temperatureĮxperimental data of tobacco dry weight were extracted from Fig. 1 (A–E) in using an open source software DataTheif (B. The following Sections 2.1.1– 2.1.2 detail the development of the field growth yield assumption distribution.Ģ.1.1. The field growth yield probability distribution was derived from an analysis of previously published literature. Harvesting time, plate & flame removal, and field growth time distributions were based on assumptions determined using working process knowledge not directly supported by reported values in public literature. Cation exchange chromatography (CEX) losses were assumed to vary according to a uniform distribution with minimum and maximum values ☑0% their mean (base case). UF/DF filtrate flux was assumed to vary by ☒5% from the base case value according to a triangular distribution.

Mechanical harvesting (P-21/GBX-104) time variability was represented by a beta distribution with minimum and maximum values based off an assumed 1–10 km/h harvester speed.

A triangular distribution was used to represent the uncertainty with field growth time before induction in P-16. The data were normalized so that the mean is 1.50 g/kg (base case expression level) and was best fit by a logistic distribution. The probability distribution profile was obtained from Werner et al. The probability distributions are not based on any existing commercial facility capability.Įxpression level variations were performed by changing the mass coefficients of “Product” and “Biomass week 6” in P-18’s RXNSEP-1, while keeping their sum constant. We used our working process knowledge to select probability distributions reflective of plant-made pharmaceutical production that one might observe at lab- and/or pilot-scale production. Assessment of assumption distributionsĪssumption distributions were primarily determined by working process knowledge supported by reports in literature. Cost breakdowns (base case + facility oversizing).Ģ.1. Techno-economic output metrics (e.g., cost breakdown by section and cost item, total capital expenditures, number of batches per year) are generated in the techno-economic modelling software using input parameter values associated with Monte Carlo-based techno-economic simulation trials that yielded the minimum, mean, and maximum values of internal rate of return after tax for the base case and facility oversizing scenarios, as described in the data file, 09. The contribution to variance of each input parameter to each forecast variable is calculated by rank correlation coefficient using Monte Carlo-based techno-economic simulation run data for the base case in which Pearson correlation coefficients are not included, the results of which are described in the data file, 08.

Forecast univariate sensitivity charts (base case). Forecast univariate sensitivity data (base case) and 07. Univariate sensitivity of the forecast variables to the input parameters is investigated using tornado plots and spider charts in the data files, 06. Forecast variable normality (base case + facility oversizing), as assessments of normality. Box plots and quantile-quantile plots are shown in the data file, 05. Statistical test results (base case + facility oversizing). The forecast variable output data are compared between the base case and facility oversizing scenarios using two-sample t-tests for evaluation of the means and Kolmogorv-Smirnov tests for evaluation of the distributions, which is summarized in the data file, 04. Analysis of techno-economic forecast variable outputs Equipment oversizing specification (base case + facility oversizing).ġ.2. The details of the equipment oversizing scenarios of the techno-economic process model to accommodate the uncertainty of production are described in the data file, 03.

#MONTE CARLO SIMULATION TOOLS CRYSTAL BALL TRIAL#

Simulation trial data (base case + facility oversizing). These assumptions distribution trial data feed into the techno-economic process model (publicly available at ) to generate the forecast variable output data in the base case scenario and facility oversizing scenarios (in which the equipment of the facility is sized larger to accommodate the uncertainty of production). The input parameter assumption distributions and associated Monte Carlo sampling-based trial data for the base case techno-economic process model are described in the data file, 0.1 Assumption distribution & trial data (base case). Generation of techno-economic data under uncertainty

Monte carlo simulation tools crystal ball