storage-based hydro
I wonder if anyone could help me how the seasonal and day-night level storage capacity of a storage based hydro-power plant could be specified (is there any good example on the TIMES documentation-maybe in the examples). Which parameters do i need to use? Do i need to create an independent storage process?



Deju Wrote:I wonder if anyone could help me how the seasonal and day-night level storage capacity of a storage based hydro-power plant could be specified

I wonder what you mean by specifying the storage capacity. If you mean defining the amount of existing capacity, you can just define it be specifying a value for the PRC_RESID or NCAP_PASTI parameter. By default, storage capacity is interpreted in terms of the amount of energy that can be stored, but if you use NCAP_AFC(output), the capacity will be the nominal max. output capacity, like with power plants. Finally, with NCAP_AFC(ACTGRP) you can define the availability for the stored energy even when also using NCAP_AFC(output), i.e. in a way you can then define both the output capacity and the storage capacity for the same process (this requires TIMES v3.8.3 or above).

Deju Wrote:Do i need to create an independent storage process?
If you want to define the storage capacity on both the seasonal and daynite levels, I would say that you need to create a separate process for the seasonal storage.

Thank you very much Antti, so i can define the usable which is 90% (of the maximum storage capacity ) during late summer and 20% early spring, is that what the NCAP_AFC(output) mean ?

Here is my data i want to define a maximum storage capacity of 10PJ and net usable 7 PJ (90-20%). The lower storage level is 20% and the maximum storage level is 90%, so isn't enough to use only the STGIN_BND input flow parameter and storage efficency to get it done?


Hmm... not exactly, but  I am not sure I follow.

As I said, if you use NCAP_AFC(output), the capacity will be the nominal max. output capacity, like with power plants.  So, in that case the capacity does NOT represent the max. amount that can be stored.

If your flow unit is PJ, maybe you should consider defining a SEASON level storage, having both input and output INFLOW with PRC_RESID = 10 (PJ), PRC_CAPACT=1, NCAP_AF(UP)=0.9 and NCAP_AF(LO)=0.2, and some storage efficiency/loss. That would be your seasonal storage part. If you create an ANNUAL level process producing INFLOW, you can define the seasonal hydro INFLOW distribution with COM_FR(INFLOW).

Then you could add a DAYNITE level storage hydro power plant consuming INFLOW (=input) and producing electricity (=output). For this DAYNITE storage you could use NCAP_AFC(output) = X, in order to define the capacity in terms of MW/GW, and the availability and capacity costs accordingly. And, as I already mentioned,  the daynite storage capacity can then be defined by using NCAP_AFC(ACTGRP).

Thank you very much Antii, i will experiment this and let you know.


Dear Antii,

Thank you very much for your help. I have been trying different scenarios but it seems I'm not able to catch the point. I'm very new to the ANSWER-TIMES tool (6.9.1 version). I managed to model the run-of-the river-hydro, wind and solar successfully, but still I'm struggling with the reservoir.

Here is what I did, I have created a separate seasonal storage process and power plant (turbine) process, linking the output of storage to the turbine with different inflow commodities (HYDS-1 and HYDS-2) inflow. For the time being I ignored the DAYNIGHT storage. I have defined the seasonal storage as you recommended last time. I'm getting power output when I'm not getting any outflow from the storage?

Storage capacity 18 PJ (annual), power plant capacity 1GW, Capacity factor 41%. Availability 90% (UP) and 20% (LO), and Power output DAY-NIGHT level. I wonder if you could show me with this real life example.


Here are the results attached:



Ok, you are using an older version of ANSWER-TIMES.  You should consider upgrading to the latest version.

I wonder why you now have two inflow commodities HYDS-1 and HYDS-2, and how they are linked to the processes?  From my earlier post you can see that I suggested a seasonal storage that would be charged and discharged by a single seasonal inflow commodity, INFLOW.

Anyway, it appears that your  current main problem is not related to storage at all. The results you provided show clearly that the power plant process E001 (which I assume to be now a normal process, according to your description?) does not have any transformation: It produces electricity but does not consume anything. But it is impossible to conclude from the results why you don't have any transformation equation for it.

Some questions that might help resolving this problem:

  • Did you define the process transformation for the process E001, and if you did, how? For example, did you define a process efficiency for it, and if so, how?
  • What is the primary group of the process E001?  (It apparently should be ELC).
  • What are the input(s) to the process E001?  Only HYDS-2?  Or both HYDS-1 and HYDS-2?
  • What are the types and timeslice levels of the commodities  HYDS-1 and HYDS-2?
Dear Antti,

Thank you very much for your help in figuring out the problem. I have declared seasonal COM_FR for HYDS-1(INFLOW), which is directed to seasonal storage process and the output of the storage is HYDS-1(INFLOW). Then i have created a separate hydro turbine process (normal process without storage) consuming HYDS-1 and output (ELC) at  DAY-NITE level with a certain annual availability factor. When this does not work, i added additional normal process which consume HYDS-1(out put of storage) and produce HYDS-2(INFLOW) which ultimately directed to turbine - this does not seem logical but i gave it a try though it does not work either.

To answer the questions: 1)Yes i have defined 97% efficency, AF(annual) 41% 2) PG is ELC 3) both have seasonal timeslice levels.

My question is why the storage output is not linked with turbine input?


Well, as I said, the foremost problem appears to be related to the E001 process, as it does not consume anything.  And so, of course the storage is not supplying anything, as there is no consumption for it!

Unfortunately you did not answer all my questions: You say that you have defined 97% efficiency, but you are not telling me how, even though I specifically asked for that information. And you did not disclose the commodity types of HYDS-1 and HYDS-2 either, which I asked for.

I am afraid that with the insufficient information provided, I am not able to help further, unless you provide the full test model for me to look at, or post screen captures showing all the process characterizations you have made. Then I would be able to see how you have actually defined the processes.

Finally, as long as you are using ANSWER-TIMES v6.9.1, you will not have the more recent storage features of TIMES available.


Dear Antii,

Thank you so much for your help. And, i'm sorry for being not able to provide you enough information so far. I've made some progress, by using a single INFLOW (HYDS) commodity as you can see on the pictures. I wonder if you have a look at on the attached screen capture of each process with their associated parameters. I hope this might be of help to figure out what is going wrong. I've also updated to ANSWER-TIMES 6.9.7.



Ok, I can see you have managed to get the process E001 working. The crucial mistake I can see in the model is that HYDS is defined as a free renewable. As such, it has no commodity balance, and therefore modeling a storage process for it would be completely meaningless.

But I did not see any indication of a supply process for HYDS; you should create such as well. I would thus suggest that you test your model after doing the following modifications:

  • Change the commodity type of HYDS from free renewable to limited renewable (this is crucial)

o        You can also remove the superfluous COM_GMAP(REG,NRG,HYDS)

  •  If you still have no supply process for HYDS, you need to create such now:

o        Output = HYDS, PCG=HYDS, no input required

o        ANNUAL level

o        Set also a reasonable upper bound ACT_BND(UP), for the annual inflow

  • I suggest to disable the second storage process (ST0002) for now, because:

o        It has no capacity bound / investment costs, and might thus have infinite storage level without any cost impact

o        Note also that it is pointless to use NCAP_AFC(ACTGRP) if you are not using NCAP_AFC(output)

Greetings ANTII

Can you tel me how to set NCAP_AFC in output?
The parameter has the following syntax:
NCAP_AFC(r,y,p,cg,tslvl) = value;
Here r=region, y=year, p=process, cg=commodity/group and tslvl=time-slice level.

To define the availability factor for the output commodity COM, you should thus specify
NCAP_AFC(r,y,p,'COM',tslvl) = X;

For this to be meaningful, the process should have a capacity. Therefore, you should also define for the process either the existing capacity, or the investment costs for new capacity, or both.

Forum Jump:

Users browsing this thread: 1 Guest(s)