3DSL v2021.0129 on 01/29/21

Bug Fixes

1. Boundary efflux rates were not properly computed when SURVEILLANCE=ON and in forecast mode.
2. If negative saturations are computed in well pair bundles for SURVEILLANCE=ON and the MODEL=FCM or LIVEOIL and these bundles are connected to effluxing open boundaries, the negative saturations are now reset to zero.  Only positive saturations can efflux from open boundaries.
3. Bug fix to read/write of .rst files for SURVEILLANCE=ON runs.

Enhancements

1. Results on a single-thread run will no longer be identical to 3DSL v2020.1125 due to:
   1. SAMG solver upgraded to 2020.0820 for improved multi-thread performance.  
   2. Performance improvements in streamline tracing and pvt calculations.
2. CHEMICAL mass rates reported at the field level are rates due to all wells in the field only and no longer included mass rates influxing or effluxing at boundaries.
3. Injector (or producer) centered off-set allocated production (or injection) is now written to the *.vectors.h5 format, so that studioSL does not need to compute these values.  Result is improved loading of per-pattern 2D plots.
4. The .wpv file is no longer written out by default as all information is now contained in the *.vectors.h5 file. If the WPV keyword is present in the OUTPUT section, this will continue to work as before and will force output of the .wpv file.