Yes it is. we are triggering a job in ECC usign this web service. and only one job can be active at a time for the others application fault is returned sayign that service is busy. Now due to this 1min delay multiple messages get collected (which may have come in every 5 secs) and all of them are sent to teh web service together instead of the actual 5 sec delay.
As this is a limitation of PI BPM that whatever time you mention in wait step is technically not the exact time the message waits due to its dependency on the SWWDHEX job, we are left with nothing but to look at reducing the frequency of the job itself. All this job does is reset a flag and usually takes 1 sec to finish.
Any suggestions on BPM redesign or anything else is most welcome.