You are correct in that the total time will first of all vary between journals and fields but also slightly over time. I and chief editor of an international journal in the environmental sciences. The time each reviewing editor spends on a paper with our journal can be broken down as follows:
- Must scan the paper upon receiving it for assigning reviewers (probably takes an hour, maybe less)
- Assign reviewers (does not take much time but can be drawn out over time when reviewers decline to review)
- Read reviews carefully and provide authors with guidelines on what to focus on (probably takes a few hours)
- Carefully check the revisions and decide on the faith of the paper (takes a couple of hours at least depending on extent of revisions)
- If the paper goes to major revisions you need to go through steps 2-4 again
- If the paper is accepted then spend maybe 15 minutes to half an hour formulating a suggestion for final decision (depends on what might need to be written)
This can be summed up for one paper as probably more than half a day. Then the question is how many papers are you requested to handle per month/year etc. You can then easily multiply by the number of papers to get a reasonable idea of the total time you will need to spend.
To add to this, you will need to act whenever a paper arrives so you are expected to be more or less on call all year. Most journals have systems for indicating when editors are away but that only works for assigning new papers, if you have started the process you will have to see it through ... and keep chasing late reviews and delayed revisions.
So although I cannot give you an idea of workload, you have some tools to figure out what will be involved. Knowing the number of papers you are expected to handle is the most important statistic to figure out. Then depending on your field, you may have a sense of whether my estimates for scanning and reading materials is reasonable. I am sure they will vary depending on discipline.