More realistic instrument deployment timings: CTD(_BGC)s, Argos, Drifters

[j-atkins]

Following on from #249

This PR:

• Changes the default CTDConfig.stationkeeping_time to 50 minutes.
  • This includes the 35 minutes assumed for ship slowing, deployment/recovery time and ship speed-up time + a middle-of-the-road 15 minutes additional time to account for the raising and lowering of the CTD (a temporary fix before we include the depth dependence properly; see comments in Realistic time for instrument deployment #249 and Refactor 'simulate_schedule` #236)
• Adds a stationkeeping_time field to both ArgoFloatConfig and DrifterConfig, defaulting to 20 minutes, to account for the ship slow-down and speed-up at deployment.
  • And carries over these additional fields to _plan logic.
• Adds a check in simulate_schedule of whether both a CTD and CTD_BGC deployment is selected at each waypoint. If so, then only the time cost for one is added, as in reality they would be the same instrument.
  • This probably indicates all the more reason to move towards CTD and CTD_BGC instruments being unified...I'll make an issue.