Let’s discuss the issue of sampling chemical products
Thread Content
Generally speaking, the sampling error is greater than the analysis error; therefore, the representativeness of the sample has a significant impact on the measurement results. For general chemical products, standards such as GB/T 6678, GB/T 6679, GB/T 6680, and GB/T 6681 are commonly used to guide sample collection, or the product standards include a dedicated chapter on sampling. I wonder how well everyone in the marine industry adheres to the standards when taking samples? We found that there are many difficulties in implementing the standards, especially when using products (raw materials) that are not produced by us. This is manifested in: sampling tools that are not suitable, difficulty in accessing the sampling sites, difficulty in achieving the required number of sampling units, difficulty in meeting the requirements for batch processing of products, and sampling methods that do not match the actual conditions. To give two examples: sampling of sulfuric acid in tank cars – For industrial sulfuric acid, which is the raw material, in accordance with the current standard GB/T534-2002, each storage tank is considered as a batch, and the sampling is carried out in line with the provisions of 2.2 in GB/T 6680-1986. And we use tank trucks, so neither batch processing nor sampling methods are suitable. If the sampling method for tank car samples specified in GB/T 6680 is used, whether samples are taken from the upper, middle, and lower parts or from the entire liquid level, it is very difficult to do so due to the large diameter of the tank cars. Generally, only samples from the upper part can be taken; however, caution is needed in winter, as sulfuric acid may crystallize, resulting in lower values for samples taken from the upper part. For example, in the case of activated carbon used as a catalyst carrier, the standard requires it to be packaged in batches of 1 ton, using 10-kilogram bags; each batch shall not contain more than 100 bags, with 14 bags constituting one sampling unit. But we arrive as raw materials, in quantities of 20 to 40 tons each time, transported by freight cars; there are 2,000 to 4,000 packs in total, representing a mixture of dozens of batch numbers from different manufacturers. How should we select them? Set your own sampling plan, but if the quality of the other party’s products varies greatly, our sampling plan will carry a higher level of risk. Do any other sea friends have similar problems? Take it out and let it dry in the sun!
It is indeed a problem. When it came to inspecting raw materials, I was lazy: I would check the qualifications of the manufacturer, and if the first few batches were of good quality and met the quality standards after several tests, I would relax the inspection process. For example, reduce the sample size as well, and select only a few key items for analysis. Three months later, the product inspection certificates provided by the other party were directly used as the basis for acceptance. However, the inspection of products before they leave the factory cannot be relaxed; if problems arise, the boss will hold someone accountable. Please point out any mistakes that were made. I’ll change it. This post was last edited by thong0421 on 2009-2-27 09:29.]
For sampling from tank trucks, there is no need to worry about the uniformity of the material, as it mixes thoroughly within the tank during transportation; however, sampling should be carried out promptly once the tank truck arrives at the facility. When it comes to sampling activated carbon, the most important factors are randomness in sampling and the quantity of samples collected. As you mentioned, the quantity is sufficient. For multiple batches of raw material, attention should be paid to ensuring random sampling – it isn’t necessary to conduct inspections on all of them.
This post was last edited by zbx on 2010-5-31 20:15. Ugh, chemical products do indeed have such problems. Our liquid products are produced in batches of 1 to 2 tons, but when a batch size of over 10 tons is required, it seems there is no better solution. If the pharmaceutical factory requires deliveries in batches, we have to process them in batches as well. There are many batches every day. Warehouse management also makes it difficult to arrange items by batch due to insufficient space.
Liquid sampling and gas sampling are both a bit better. It is very difficult to ensure the representativeness of solid samples. In our company, raw materials undergo two levels of inspection upon arrival at the factory. For materials such as calcium carbide, coal, fluorite powder, and phosphate rock, it often happens that the samples pass the tests conducted by the head office, yet the production workshop insists that they are Unqualified. Samples were kept for reference, and the results were indeed correct. Several sampling methods have been established, their implementation is very strict and supervision is thorough, yet disputes still arise. It only makes the sampling method more detailed and more complicated. Sampling is often carried out in the three stages of production, inspection, and warehousing, in accordance with national standards or product sampling requirements; yet it can still be difficult to reach a consensus on the results.
The quality of sampling has a direct impact on the accuracy of the overall results. For the methanol water solubility test in our unit, it is not easy for those without sufficient experience to determine whether the product is of top quality or first-class quality; the 1+3 and 1+9 tests also make it difficult to make such judgments.