Emerson Exchange 365

We redesigned a 'clean in place' (CIP) operation and now during simulation testing the recipe won't load in batch operator interface. The load_failed message is: "Not enough storage is available to complete this operation".

Checked both disk space and freemem and those are sufficient. Also checked BOL and guardian KBA's, but couldn't find any reference to this problem and the batch limits listed in BOL aren't reached by far. On other units we already have large (CIP) operations that run fine. The largest OP now has 63 phases and 31 transitions. The new OP has 73 phases and 35 transitions, so slightly more. (In the past we had an OP with 118 phases and 160 transitions which ran fine for months continuously.) SFC logic is all valid and verify & download runs ok. All PLM's are assigned to the correct unit-module and downloaded. (We use PLM's instead of phases classes, because we need to run over the 10 parallel phases on 1 unit.) Transition expressions all double-checked. Other operations on this unit load and run fine.

What we tried is to make a copy of the new OP and start removing phases and transitions from the bottom up, then download and try to load. After the number of phases went down to 50, the recipe loaded without error. When we started removing phases and transitions from the top down, we ended up with around 30 phases before it would load in BOI. So there is a problem somewhere below the middle, but it's not obvious what it is. Solution could be to split the OP into 2 parts, but we want to avoid this.

What I also noticed is that the compiled size of this OP (in the download folder) is much larger than the other, 53kb vs 40kb and the largest OP in the system. The number of lines (single & parallel) account for a lot of the compiled data and the new OP seems to have many more lines than the other OP, so maybe there is a limit there?

Anyone else had similar problems with loading large recipes and know of a solution? Version is 10.3.1 with all hotfixes installed.