Releases: DTOcean/dtocean-maintenance
Releases · DTOcean/dtocean-maintenance
Release list
Version 3.0.0 Release
This major release improves performance by utilizing version 3 of dtocean-logistics and optimising the use of pandas DataFramaes. Integration with version 3 of dtocean-reliability has also improved component failure calculations, detection of string failures that affect multiple devices and inclusion of reliability k-factors.
Detailed changes are as follows:
Added
- Add logging to report selected ports.
- Added start_logging command to configure and start the logger.
Changed
- Changed om_logistics_main function to Logistics class to reduce repetitive initialisation.
- Integrated changes to dtocean-reliability module, with the following impacts:
- Storing a single Network object reduces repetitive initialisation.
- Improved identification of curtailed devices upon string failure.
- Integration of k-factors into reliability calculations.
- The reliability input interface (RAM_Param) has changed.
Removed
- Removed excessive use of DataFrame appending using loc.
- Removed costly calls to DataFrame.sort_values.
- Removed unused ComponentIDLogistic variable.
- Removed filtering from poisson_process function.
Fixed
- Fixed case where no calendar based events are scheduled.
Version 2.0.0 Release
This release coincides with the release of DTOcean 2.0.
The following changes were made:
Added
- Added option to logistics algorithm to optimise to minimum start delay. This
is utilised for the corrective maintenance strategy. - Added post calculation analysis classes and methods to the static module for
energy, availability, logistics and financials. - Added statistical method for generating multiple data points calculated with
varying random seeds for sub-system failures. The results of all data points
are now returned for external analysis. - The OPEX component of LCOE is now calculated and returned.
- Vessel names are now included in output events tables.
- Added startProjectDate input to collect the actual start date of the project
so energy and OPEX tables start from year 0. - Added curtailDevices input to trigger option for devices to be shut down
indefinitely on failure (not tested). - Added numberOfSimulations input to collect the number of data points to be
used for statistics. - Added NumberOfParallelActions input to set how many maintenance actions
can be completed by one vessel (default to 10). - Added correctivePrepTime input to allow the amount of time required for
preparation of a corrective maintenance event to be set (defaults to 48
hours). - Allowed exclusion of sub-systems from calendar or corrective maintenance if
they are given maintenance intervals or health thresholds of zero or less.
Changed
- Changed package name from dtocean-operations to dtocean-maintenance.
- Renamed modules.
- Ensured failure rate is always read from the dtocean-reliability module.
- Ensured ports and weather windows are always read from the dtocean-logistics
module. - Refactored LCOE_Calculator.__postCalculation using new static post-analysis
functions and improved outputs. - The power_prod_perD input is now a dictionary, containing the device names as
keys, rather than a numpy array. - A RuntimeError is now raised if no weather windows can be found for an
operation.
Removed
- Removed whichOptim, checkNoSolutionWP6Files and integrateSelectPort inputs
as no longer functional.
Fixed
- Fixed bug in distinguishing between reliability data of moorings and
umbilical cables. - Fixed bug in retrieving reliability data for multi-hub layouts.
- Fixed bug which occurred if no corrective maintenance events where simulated.
- Fixed bug with port selection if no operations are requested.
- Fixed bug which occurred if the number of operations was divisible by the
number of parallel ship operations. - Fixed start delay from corrective maintenance operations when previous
operation not completed before the scheduled start date. - Fixed labour costs not being shared when undertaking parallel operations.
- The corrective maintenance table is now correctly sorted after addition of
downtime (which extends sub-system lifetime). - Fixed scheduling of calendar based maintenance to respect the allowed months
of operations and ensure that the first operation occurs within the given
range. - The sub-system MTTF is now used to determine how long a sub-system with
calendar maintenance will exhibit perfect reliability from deployment and
after maintenance. - Added waiting time to operation duration calculations, which was missing.
- Reorder check on the impact of calendar maintenance on corrective maintenance
events and the shift their operation request dates. - Fixed check for day after which failures can occur for sub-systems with
calendar based maintenance, when no maintenance actions are scheduled. - Fixed double counting of umbilical cable reliability data if both electrical
and moorings modules inputs are given.