|
本帖最后由 张鹏326 于 2020-7-14 11:38 编辑
2.4 What can we do?
2.4 我们能做什么?
We should change the definition of solution.
我们应该更改解决方案的定义。
The oil production depending on time is not a one graph, but a set of graphs, showing the uncertainty of this solution. All the graphs in this band are solutions.
取决于时间的采油量不是一个图表,而是一组图表,显示了该解决方案的不确定性。该区间内的所有图形均为解决方案。
Sensitivity analysis
敏感性分析
We should check the solution sensitivity to changes in calculation parameters. In other words, we should check
我们应该检查解决方案对计算参数变化的敏感性。换句话说,我们应该检查
- Sensitivity to time step size limit,
- 对时间步长大小限制的敏感性,
- Sensitivity to accuracy of linear and non-linear problem increase,
- 线性和非线性问题增长的精度的敏感性,
- Sensitivity to number of cores.
- 对内核数量的敏感性。
We should check the solution sensitivity to changes in model parameters. There are many of them. Check at least those that may be automated easily:
我们应该检查解决方案对模型参数变化的敏感性。(参数)有很多。至少检查那些可以轻松实现自动化的功能:
- Sensitivity to relative permeability end points,
- 对相对渗透率端点的敏感性,
- Sensitivity to layer permeability increase,
- 对地层渗透率增加的敏感性,
- Sensitivity to GOC and WOC.
- 对GOC和WOC的敏感性。
(注:通常,我们没有时间和精力去分析这些敏感性,一般也很少遇到这种敏感性的问题;另外,很少有人自曝其短,就像不会有哪个油藏工程师经常将“数值模拟是近似手段、具有不确定性”挂在嘴边,虽然这在一定程度上就是事实。--qcy)
|
|