Does anyone know if the rolling average as described in ACI 301 sections 1.6.6 and 1.6.7 apply when the same mix design was being produced at two different concrete plants?

We molded 7 sets of cylinders for a large elevated deck pour, and the 5th set of cylinders we molded came from a truck that was from one plant, and all of the other sets of cylinders we molded came from another plant.

The one set that we made from a different plant was the only one that did not meet the specified f'c.

The intent in ACI 301 and ACI 318 is "a single facility", not multiple facilities producing the same mix design.  While some useful information can be gained from included multiple sites, this is not one of them.

In computing the moving average, use only those results from a single batch plant.

and the same mix design (and components of the mix).

Did the same technician take the sample and make all the sets of cylinders?

Were all handled and cured in a similar manner.

What is your climate? Different plants can have different aggregate temperatures depending on the storage method and rate of use if you are in a cold climate. This will have an effect of 7 day or quicker breaks.

do you have access to print-outs of the individual batches - most good plant systems can retreive the information. This can be helpful to determine if the same materials were used.

When covering a large pour where the mix is coming from two different plants is it incorrect to intentionally make sets only from one plant?

What if plant A breaks down during a pour and a switch is made to plant B?  

It in not correct to take specimens from only one plant.  After all, the tests are intended to represent the mix design being used.

In evaluating a moving average, it is appropriate to use the data from each plant, but separately.  Do not mix data from one plant to another for a moving average computation.

Ron, interesting...I would have said that as long as the mix design and all the materials are from the same source (individual class of concrete), then yes, it would be considered by the running average of three.

You may wish to monitor trends, one plant vs. another, but the mix is considered the same mix if the mix ID and ingredients are the same.

boffintech...in theory, the mix is the same, I agree.  However, the moving average is a statistical concept that is intended to reflect the capability of the mix and the capability of that plant to consistently produce the mix with similar results.  If you use the data from two plants, you have two variables you're trying to evaluate from a single data set.

Even though the mix "recipe" might be the same and being produced by two different plants of the same company, there are still variables between the plants. They have different scales, their storage/conveyances are often different, they use water from two different sources, and variations in production rate might mean that material sources/timing may change enough to affect the mix consistency.

I can envision doing both individual plant and combined, properly labeled.