[猫三Iridio]

有四个实验：1、纯注气，2、纯注水，3、水气交替注入（1：1），4、先注水，等采收率达到注水开采的极限了，然后进行水气交替注入（1：1，共有九个注气段）。前面四个实验已经拟合完毕，效果较好（因为相渗也是做过试验的，所以拟合起来比较轻松）。但第四个实验，前面注水阶段和第一第二注气阶段均拟合较好，但从第三注气段开始，采收率便急速上升，为什么？（Ps：在拟合的时候，发现注气实验（包括前面几个），气窜时间可以拟合的很好，上升段也还行，但是实验的汽油比是不断上升，而Eclipse模型的汽油比却是上升到一定水平后保持不变，这种情况正常吗？（已经试了加大注气量，效果一样））。本人接触组分模型不到一个月，道行很浅，请高手指教，万分感谢。

附加文件

•  图片4.jpg (42.27K)
  下载次数: 12
•  图片5.jpg (20.44K)
  下载次数: 8
•  图片6.jpg (21.81K)
  下载次数: 4

[猫三Iridio]

怎么没人回复啊，高手都跑哪去了？

[LXH536]

1.不错的题目，把你的Krow and Krog 发上来
2.注入的什么气？干气？CO2?.....
3.涉及到混响没有？从你的采收率来看，多半是混相了，那你就要确保你的Model 也 混相....， 检查EOS 和model 压力.....

[猫三Iridio]

引用框（LXH536 ＠ 10-17-2011 - 09:39）

1.不错的题目，把你的Krow and Krog 发上来
2.注入的什么气？干气？CO2?.....
3.涉及到混响没有？从你的采收率来看，多半是混相了，那你就要确保你的Model 也 混相....， 检查EOS 和model 压力.....

相渗如附件所示，注入气是湿气，但PVTi上显示的混相压力为49MPa，而实验压力在42MPa左右。刚点上Miscible Gas Flood这个选项，里面增加了设定表面张力的内容，如果要调这个的话该怎么调。谢谢您的答复

附加文件

•  图片1.png (3.17K)
  下载次数: 19
•  图片2.jpg (57.33K)
  下载次数: 18

[LXH536]

1, Krog looks like abnormal.
2,The critical gas saturation can be reduced from 0.08 to 0
3,meanwhile，decrease the Kro and increase Krg in the Krog table.

have a try this one！
0 0.00001 1
0.05 0.001 0.65
0.082 0.009 0.52
0.123 0.016 0.41
0.164 0.052 0.25
0.205 0.09 0.15
0.246 0.15 0.1
0.287 0.19 0.05
0.328 0.28 0.03
0.34 0.32 0.0017
0.369 0.4 0.000001
0.4 0.5 0
0.5 1 0

[猫三Iridio]

引用框（LXH536 ＠ 10-17-2011 - 17:44）

1, Krog looks like abnormal.
2,The critical gas saturation can be reduced from 0.08 to 0
3,meanwhile，decrease the Kro and increase Krg in the Krog table.

have a try this one！
0 0.00001 1
0.05 0.001 0.65
0.082 0.009 0.52
0.123 0.016 0.41
0.164 0.052 0.25
0.205 0.09 0.15
0.246 0.15 0.1
0.287 0.19 0.05
0.328 0.28 0.03
0.34 0.32 0.0017
0.369 0.4 0.000001
0.4 0.5 0
0.5 1 0

谢了，呵呵，之前走过许多路，调相渗已经调过了，总体作用不大。

[猫三Iridio]

引用框（LXH536 ＠ 10-17-2011 - 17:44）

1, Krog looks like abnormal.
2,The critical gas saturation can be reduced from 0.08 to 0
3,meanwhile，decrease the Kro and increase Krg in the Krog table.

have a try this one！
0 0.00001 1
0.05 0.001 0.65
0.082 0.009 0.52
0.123 0.016 0.41
0.164 0.052 0.25
0.205 0.09 0.15
0.246 0.15 0.1
0.287 0.19 0.05
0.328 0.28 0.03
0.34 0.32 0.0017
0.369 0.4 0.000001
0.4 0.5 0
0.5 1 0

而且相渗是对的，使用SWFN、SGFN和SOF3这类型相渗计算SOMGAS时，临界油、水饱和度正常

[LXH536]

1,May be the Krog calibration can't work, however the Krog curve looks unreasonable.
2,To increase the GOR when the injection gas breakthroug, focus on the miscible pressure.Review the 3D model to check whether the the gas breakthroug when 1 PV gas was injected .
3,Meanwhile,look at the remain So,the remain So in your model may be zero (the oil recovery nearly 90%).But the lab data show that the injection in immisciable flood.So you may need calibrate your EOS or initial Zi friction!

[猫三Iridio]

引用框（LXH536 ＠ 10-18-2011 - 08:53）

1,May be the Krog calibration can't work, however the Krog curve looks unreasonable.
2,To increase the GOR when the injection gas breakthroug, focus on the miscible pressure.Review the 3D model to check whether the the gas breakthroug when 1 PV gas was injected .
3,Meanwhile,look at the remain So,the remain So in your model may be zero (the oil recovery nearly 90%).But the lab data show that the injection in immisciable flood.So you may need calibrate your EOS or initial Zi friction!

你说的很对，一共18个网格，到了末期已经有十几个网格So=0，我再重做一下PVTi拟合看看。

[猫三Iridio]

引用框（猫三Iridio ＠ 10-18-2011 - 09:12）

引用框（LXH536 ＠ 10-18-2011 - 08:53）

1,May be the Krog calibration can't work, however the Krog curve looks unreasonable.
2,To increase the GOR when the injection gas breakthroug, focus on the miscible pressure.Review the 3D model to check whether the the gas breakthroug when 1 PV gas was injected .
3,Meanwhile,look at the remain So,the remain So in your model may be zero (the oil recovery nearly 90%).But the lab data show that the injection in immisciable flood.So you may need calibrate your EOS or initial Zi friction!

你说的很对，一共18个网格，到了末期已经有十几个网格So=0，我再重做一下PVTi拟合看看。

气已经突破了

[LXH536]

Three proposal solutions:

1,Reduce the model pressure below the misciable pressure
2,Increase C1 or decrease C3-C5 friction of injection gas.
3,Increase BIC

BTW,if you can provide you EOS and Zi of injection gas. we can give you suggestions directly.

[猫三Iridio]

引用框（LXH536 ＠ 10-19-2011 - 09:02）

Three proposal solutions:

1,Reduce the model pressure below the misciable pressure
2,Increase C1 or decrease C3-C5 friction of injection gas.
3,Increase BIC

BTW,if you can provide you EOS and Zi of injection gas. we can give you suggestions directly.

油样为临界挥发油，原始地层压力48MPa，泡点压力为47MPa，实验测的混相压力为40MPa，但是显然实验测的最低混相压力是原始地层油失去一部分轻组分之后测定的（因为40MPa已经远低于泡点压力）。我之前说的最低混相压力为49MPa是软件测的混相压力，他必须高于泡点压力。建立的长岩心模型，原始压力为48MPa，但有一段天然能量开采阶段，之后压力下降至42MPa（低于泡点压力），所以我猜这就是注入气样能混相的原因吧，有没有这种可能?

[猫三Iridio]

重新做了一下PVT拟合，其他实验的拟合效果好多了，但是第四个实验还是一样，求解释啊。附件是第一、二个实验拟合效果

附加文件

•  图片2.jpg (21.28K)
  下载次数: 10
•  图片1.jpg (29.39K)
  下载次数: 7

[LXH536]

The GOR of the first model that pure gas was injected was also not matched after 2 PV injection.Did you try the 3 solutions that mentioned before? I think these solutions should work.

[猫三Iridio]

引用框（LXH536 ＠ 10-20-2011 - 09:15）

The GOR of the first model that pure gas was injected was also not matched after 2 PV injection.Did you try the 3 solutions that mentioned before? I think these solutions should work.

同第三个实验对比，发现在相同的注入气量的情况下，第四个实验的采收率降低了近15%，怀疑实验有问题。有没有可能是长期注水导致的水锁效应？呵呵，谢谢您的帮助。

[猫三Iridio]

引用框（LXH536 ＠ 10-20-2011 - 09:15）

The GOR of the first model that pure gas was injected was also not matched after 2 PV injection.Did you try the 3 solutions that mentioned before? I think these solutions should work.

关于后期的汽油比，因为没给出具体的产油量数据，所以没法验证我的猜测。我的猜测是，当时的产油量已经很小，微小的产油量误差也有可能导致汽油比测量的不精确。

[LXH536]

Check the Gas rate and oil rate? why not?