What is the fluid?

I'm not sure what you mean by 'more efficient', I've seen both approaches used depending on the fluid.  The bayonet is the simpliest but the heat transfer you get is going to be less than with an external plate and frame that is pumped.  That means a bigger bayonet heat exchanger but with an external heat exchanger you have the additional costs for the pump, piping, power, etc.  Then you have the costs to run conduit and cable for the motor, an MCC, etc.

Cleaning of the external, if it's needed, isn't that difficult.  To clean the internal, if it does foul, means taking the tank out of service.  There are lots of internal tank heaters so this is just a point to consider.  Is this a clean or dirty service?

With a 60' tank if you want good heating, I think I would go with an external pumped exchanger with a jet mix nozzle to circulate the tanks contents rather than relying on natural convection.  

Although easy to dismantle, bayonet heaters (even finned) may not be so effective in heat transfer when compared with forced convection types of heat exchangers. Some more comments/questions:

• The heating duty, as mentioned by s0eebuch, would
   be in the range of 2.4 million Btu/h. It appears that
   for a regular oil in a tank almost full, this heat input
   would mean a temperature increase of about 2-2.5 ^oC/h.
• Does the tank already have the nozzles for a
   bayonet-type heater ? It may have, on the other hand,
   connections for external circulation and return through
   a mixing jet.
• As TD2K asked: what is the fluid being heated ?
• Is it possible to scorch it by high heat fluxes ?  
   If this is the expected situation, externally-controlled
   heating may be the right approach.
• What are the required fluid temperatures ?
• What are the weather conditions to which the tank
  is exposed ?
• Is the tank thermally insulated ?
• Are you planning to alternatively use a "suction
   heater" thus probably saving energy ?

Kindly enlarge on the basic information to enable a better understanding of the issue in hand.

Thanks for the responses.  To elaborate, the tank is an oil/water gravity separator tank.  The water, oil, and emulsion (interface) levels are unknown and highly variable.  The bayonet heater was attractive because it did not disturb the tank contents.  Since the idea is to aid in the oil separation from water by the addition of heat, I am concerned that the external heat exchanger will negate any benefits of the heat input by creating turbulance and re-emulsifying (is that even a word?) the contents of the tank.  The tank sits on the water-front, is un-insulated, the ambient temperature (min.) is 19 deg. F. and the average windspeed is 7.5 mph.

Thanks again for the assistance.

Some more questions:

• Have you estimated the temperature gain from the   
  proposed steam consumption ?
• Have you made a lab-bench check to see whether
  this temperature rise would suffice ?
• Have you estimated the heat losses of the tank to
  the suroundings ?
• Wouldn't it be more efficient to use an emulsion
  breaking chemical ?