I have to plan a project in which for a given task and given resource, workload changes from one month to another. How I can plan this in MS project? It is very critical for me to monitor resource loading per month.
The description of the problem suggests the work is "level of effort" and there is likely both normal variability as well as seasonal variability defined by calendar months. Scheduling this in a tool, such as MSProject, is similar from a risk perspective as any type of project task. In this case, the tasks repeats self and has a start and finish based on a monthly cadence versus a task that does not repeat itself necessarily and has a start and finish based on the work and its natural completion.
Both types of tasks have both aleatory and epistemic variability and both require choosing a planning value somewhere in the probabilistic range you estimate for that task. For example, if I was estimating a discreet task, I might arrive at a work range of 200 hours (best case), 450 hours (most likely), and 600+ hours (worst case); duration of 10 days (best case), 15 days (most likely), and 25+ days (worst case); and cost of $20K (best case), $45K (most likely), and $60K (worst case). The schedule requires me to choose somewhere in these ranges where I want to target my performance based on the risk I choose to assume for that piece of work. If I am uncertain or risk averse, I'd choose on the fatter side of the range; vice versa if I have higher confidence.
The same logic would apply for a LOE task. You could either choose to use a planning value that you would load in the schedule for all the months, i.e., each month would have the same hours / cost / duration planning values and you simply manage the variances from a cumulative perspective, or you would apply some seasonal logic to the planning values if you knew when the work would climb or descend in op tempo based on the seasonality of demand. Choose the planning value, load it in the tool, and then manage your variances as you progress in time.