Here in OR we are allowed to pinch a defined amount from any batch (flower/edible/extract) for the excuse listed as “In House QC”, which we believe is code for “I smoked it”. Or at least that’s how the folks I’m with are going
However, we’re not allowed to consume those samples on site.
How do I know my quality control technicians are doing their assigned work?
How do I collect the data? What sort of training/qualifications do my testers need? How do I make distribution of test samples fair, while still insuring data integrity?
No, I’m not kidding. We’re allowed to do this, and it’s useful information. I need an SOP. You do to…
Make a questionnaire for the volunteers to fill out. Collet a large group of people 20 or so, there consumption habits should all be once dailys. This will allow you to get data from someone who achieves baseline canaboid levels regularly.
Give half of them your product, give the other half some other stuff. Record who gets what, but don’t tell them. Collect your questionnaire sheets and see what the people had to say. The next day reverse the study and compair the data. You’ll defiantly see trends in the data of you make a spreadsheet graph.
Pharma usually does visual inspection of every single vial of finished product with high light and a white background as well as a black background.
There are acceptance criteria for quality based on visual cues. At this time you can also do qc for the containers to ensure they’re well sealed and everything. You create a %bad material and if you fail a certain percent above that you fail the lot.
we’ve got the questionnaire for our flower. it could use work.
once daily is rough. right after work (at home) is the best compromise we’ve found. at wakeup was problematic.
I did have some problems before figuring out tolerances for individual testers (for tinctures/pills/edibles), but once establish, handing out only single servings helped avoid performance issues the next day.
with flowers we blind the strain, same when testing strain specific tinctures. For dabs, I’ve been going with un-blinded. Always making sure to get acknowledgement of specific dosing for anything other than flower or dabs.
I’m not sure the rules allow me to pinch enough to achieve 20 testers. I use a rotating crew of about 10. Problem is some testers are better at it than others. Which has created tension when some testers are provided samples more often than others (at least for flowers, which is not my department). For my extract testers, they had to prove a palate, and appropriate equipment. I didn’t get around to a written survey for extracts before spinning production down to await permitting.
To be honest, I suspect the tabulation of flower data has been dropped in my absence. I don’t know if our new kitchen witch has formalized his process yet.
Using the “trade sample” excuse to get extract (5g) into the kitchen gives him the ability to do product R&D. the data gathering requirement would then be the same. very few of our testers are going to notice the effects of the 5mg cannabinoid content, but I do have a couple or three.
I think to develop an SOP you have to start getting away from testing that is prone to subjectivity and find analytical metrics that can be measured to demonstrate consistency of product. Anything besides visual and scent testing wouldn’t be useable data on paper, although it could be helpful in helping you make personal quality decisions.
What specifically do you want to include in an SOP for organoleptic QC?
(Keep in mind, this should be the last step after determining analytical metrics)
I don’t disagree in general, and I definitely ALSO need SOP’s that define how to visually monitor product consistency and quality during the processing and packaging phases. and specify complete or spot checking after the fact as a separate task.
The first data point that comes to mind that I can’t get at any other way would be “activation time”. That might better be seen as a product development rather than a QC issue.
The folks who implemented the flower “QC” program looked into it as a way of rewarding trimmers and improving moral. Not what OLCC had in mind.
I got there because my GC can’t tell THC from THCA without nasty derivatizing agents, so I went with a bio-assay. The industry in general insisted that we needed to be able to sample our wares to produce/market them. I argued in public meetings that this meant we should also be allowed to sample onsite, to prevent diversion. didn’t fly
State calls it QC, so I’m trying to go with the spirit of the law…
I think there is valuable (quality based) information to be had from these bio-assays, even if they are subjective.
Munchies, couchlock, “energized”, cotton mouth, appetite suppression are all descriptors I get consistently in a strain specific manner. I can proxy the appetite suppression on the GC (THCV), and probably the couch lock (CBN)**, not so the others.
I have to use a 3rd party lab to get terpene profiles atm, so taste and smell are important metrics as well.
** SRI actually reports their machines as loosing the ability to detect CBN over time. I think it’s a specific column. Not the one I’m running.
Edit: there is also good evidence that myrcene is the primary compound responsible fo “couchlock”. CBN is merely “sedative” (?!?)
Bumping an old thread, but I would recommend having your colleagues go through the Trichome Institute’s interprening ladder: https://trichomeinstitute.com/interpening
Give me a call to discuss the " derivitizing reagent". Its not all that bad and using it only adds 2 minutes and 10 cents to the analysis.
Also, you can test for the terps with your GC really easily and at no extra cost, just a 6 minute longer analysis time.
Hugh Goldsmith
SRI
In the extraction operation I oversaw I devised a standardized form I would issue to my technicians as well as people on the cultivation side for those In House QC samples. The user would give feedback by answering questions on the form covering hardware (cartridge performance) and flavor and effect.
For internal QC, I ran HPLC cannabinoid analysis on: milled flower for extraction, adsorbents used inline extraction, crude unwinterized oil, winterized oil, and distillate. I tracked cannabinoids from flower to finished oil for a given extract batch.
We also started doing viscosity testing to try and quantify and standardize viscosity of finished product for vape pens.
We also took measurements of water content in ethanol and after a certain point would relegate ethanol lots to cleaning and non production uses
For cartridge filling, the operator would fill a form which specified the filling tolerance based on cartridge sizes. They would take net weight of the first filled 10 cartridges, and if all were in the specification, they would then take every 10th net weight.
845‐025‐1360
Quality Control Samples
Only producers and processors may provide quality control samples to their own employees. Quality control samples are only for internal quality control and product development purposes of a licensee’s own items. The items cannot be consumed on the licensed premises and cannot be resold to another licensee or consumer. Below are the limits on products used for quality control samples within a business.
Producer:
28 grams of usable marijuana per harvest lot Processor:
5 grams of cannabinoid concentrates or extracts per process lot
12 individual units of sale per process lot for other cannabinoid products
How to enter a quality control sample in CTS
Quality control samples should be recorded via an adjustment with the reason of “In‐House Quality Control.”
