Calculation of pumping energy at splitters

Hi,

I tried to calculate the total pumping energy with the cost block. I want to provide the pumping energy of the sludge waste pump which I would get from the "two flow splitter" block. However, the only two output interface I can find there are Q_Op1 and Q_op2. Why is the Pumping Energy not visible?

Best regards,

Youri

Hi Youri,
correct: the one you're using (default model for the splitter block) is meant to be coupled to the true pump block - which expects the "desired" flow (i.e. what the process needs). You should then be able to see whether or not, based on the pump + system characteristics, a single pump is able to deliver that flow; or if you need more than one pump.
With this combination of blocks, the energy / power is calculated in the pump block which can then be connected to the cost block.

However, the splitter block allows for second model (the 'old', very basic model of WEST) which, as you expect, will expose the pumping energy and power.
In fact, it's safer to use the power to compute the costs rather than energy (you will find the corresponding term in the cost block) because you could use variable tariffs (the tariff is now a manipulated variable!).

This is getting quite complicated. But OK, if that is the way that you want to go with the standard model library.

By the way, why is pumping power safer to use than pumping energy?

[quote author=Youri Amerlinck link=topic=20566.msg26517#msg26517 date=1498828548]
This is getting quite complicated. But OK, if that is the way that you want to go with the standard model library
[/quote]
Do you find it complicated?
Would you be able to suggest a more convenient solution?

[quote author=Youri Amerlinck link=topic=20566.msg26517#msg26517 date=1498828548]
By the way, why is pumping power safer to use than pumping energy?
[/quote]
Because the integral is calculated within the Cost block, including the possibly time-dependent tariff. If you would pass the energy to the Cost block, the tariff would be outside the integral, thus only work for constant tariffs.

[quote author=Enrico Remigi link=topic=20566.msg26518#msg26518 date=1498857748]
[quote author=Youri Amerlinck link=topic=20566.msg26517#msg26517 date=1498828548]
This is getting quite complicated. But OK, if that is the way that you want to go with the standard model library
[/quote]
Do you find it complicated?
Would you be able to suggest a more convenient solution?
[/quote]

Yes, I do find it complicated. The fact that you need to select between two types of splitters is complicated to explain. And I actually have no idea what the 'true pump block' is. I didn't see it either but I probably looked over it because I was not looking for it.

In my opinion, for what it is worth, you should develop a model library depending on its use. Doing a generic dynamic modelling study evaluating the plant performance is another thing than doing a detailed energetic study (including pump costs, etc.). So, keep it as simple as possible.

If you want to keep everyting in one model library I see two options.
1. Take the most basic model as the default one. If people want to calculate with the true pump block they can change the model in the splitters.
2. make one splitter model that provides all the necessary outputs.

You're actually touching upon very pertinent issues here ...
[quote author=Youri Amerlinck link=topic=20566.msg26524#msg26524 date=1499110157]
The fact that you need to select between two types of splitters is complicated to explain.
[/quote]
Yes and no ... it may be difficult to explain, but, on the other hand, a user may find annoying to have to replace an icon with another icon of the same type in order to switch model. Although, it's also true, now one can simply drag an icon over an existing icon and, if terminals are compatible, all the connections will be maintained.

[quote author=Youri Amerlinck link=topic=20566.msg26524#msg26524 date=1499110157]
And I actually have no idea what the 'true pump block' is.
[/quote]
Yes, there's a new Palette in the Block Library pane called "Pumps".

[quote author=Youri Amerlinck link=topic=20566.msg26524#msg26524 date=1499110157]
In my opinion, for what it is worth, you should develop a model library depending on its use. Doing a generic dynamic modelling study evaluating the plant performance is another thing than doing a detailed energetic study (including pump costs, etc.). So, keep it as simple as possible.
[/quote]
I tend to agree with this suggestion.
In fact, we've been thinking about something similar lately, but indeed: we could 'package' sub-sets of models into libraries aimed at different types of modelling goals ...

[quote author=Enrico Remigi link=topic=20566.msg26527#msg26527 date=1499154749]
Although, it's also true, now one can simply drag an icon over an existing icon and, if terminals are compatible, all the connections will be maintained.
[/quote]

Cool, I didn't know that you already implemented this.