Don't forget the effects of fouling on the HX performance.  This will require higher steam pressure to increase the LMTD and maintain the heat transfer than when the exchanger is new and clean.  

It is best to avoid choked (sonic) flow across a control valve where possible.  For steam (k=1.3) the critical pressure ratio for choked flow is 1.83:1.  In your application, this translates to keeping the steam pressure on the shell above 75 PSIG.

If the heat exchanger design is already finalized, then it will require a certain minimum pressure to meet your heat load requirement.  You really are only free to play with the steam pressure on the shell if you are still in the design stage for the heat exchanger.  A higher pressure on the shell will result in a smaller cheaper exchanger.

You are correct in stating that the condensate line can be rather large when high pressure steam is used.  When sized for the flash steam velocity, it often ends up being larger diameter than the high pressure steam supply line.

Thanks for the info butelja. Good point concerning the choked flow for the valve! Your also right about a smaller HX due to the higher steam pressure but the heat of evaporation is higher 969 BTU/# @ 15 psi vs 902 BTU/# @ 75 psi which causes the condensate liquid being evacuated to carry more heat 277 BTU/# @ 75 psi vs 181 BTU/# @ 15 psi. The result is a bit smaller HX but condensate piping and trap which is a lot larger. There is a trade off.

Forgive my ignorance, but what does LMDT stand for?

Yes the system is designed and installed, but I always go back and evaluate how it is working. Its easy when you can trend the variables with the DCS.

My findings are:
(1) - As you mentioned, I would take more pressure drop across the valve.
(2) - The design numbers, which were given to me for sizing, were coservative.
  
The above factors contribute to a average control valve opening of 25% instead of 50%-60% which I would like to see.

LMTD stands for Log Mean Temperature Difference.

As the temperature control valve (TCV) throttles-in on a reduced load, the pressure drop across the valve increases. You get the exactly same effect as you would with a pressure reducing valve (PRV). This is why it is important to pipe the heat exchanger such that condensate can drain by gravity. A vacuum breaker is also needed to permit condensate to drain when the TCV closes completely, and a vacuum can form in the HX. The trap needs to be sized with this greatly reduced pressure drop in mind as well. Under certain load conditions, you sometimes only have 1 or 2 PSI differential across the trap.

Consider installing a flow switch in the liquid flow, tied back to the TCV. I've seen pumps fail, or get shut off, where the sensor for the TCV is in the outlet piping from the HX. The liquid surrounding the sensor is cool enough that it keeps calling for steam. Meanwhile, the liquid temp inside the HX is off the scale.

Remember to install a properly sized & rated relief valve on the liquid side of the HX, if it can be isolated, or is in a closed system. Steam valves leaking-by, or a control failure can have energy being added to a system with no load to take it away. This can result in HUGE pressures. This is a particular problem with the increased use of backflow prevention devices. The water side used to be able to expand back into the plant or municipal water systems. Now it can't. I've seen HXs fail because of this.