Re: benefit function of dispatchable load

[Muhammd Ali Mohseni]

Dear Samuel,

thank you very much, now I understand what happen

 With respect,

  Ali

On 9/15/16, Samuel Perkin <samu...@landsnet.is> wrote:
> Dear Muhammd,
>
> The coefficients of the cost function should be positive (see Figure 6-5 in
> the Matpower manual), and as dispatchable load is modelled as negative
> generation your cost value will be negative (i.e. a benefit). The OPF aims
> to minimize costs (or maximise negative costs) the model will now correctly
> aim to increase load as much as is economically feasible.
>
> In short, what you have now is correct!
>
> Kind regards,
>
> Samuel
>
> -Original Message-
> From: bounce-120786918-76238...@list.cornell.edu
> [mailto:bounce-120786918-76238...@list.cornell.edu] On Behalf Of Muhammd Ali
> Mohseni
> Sent: 14. september 2016 17:19
> To: MATPOWER discussion forum <matpowe...@list.cornell.edu>
> Subject: Re: benefit function of dispatchable load
>
> Dear Samuel,
>
> I don't have any other problem. It is interesting when I change the benefit
> function coefficient to positive like this:
> mpc.gencost = [
>  .
>  2   0   0   3   0.5*0.015  13.00;
>  ]
> there isn't any warning and DC Opf converge!!! now results is true???!!!
> hence the power associated with dispatchable Loads is negative but first
> part of cost function is positive(0.5*0.015*P^2>0)!! Is the final value
> of Opf correct???
>
> With respect,
>
>  Ali
>
> On 9/14/16, Samuel Perkin <samu...@landsnet.is> wrote:
>> Dear Muhammd,
>>
>> Your error message is not really an error, but is just a warning
>> statement that you shouldn't need to react to. It simply means that
>> the optimisation algorithm doesn't require an initial guess of the
>> solution, and therefore any initial guess is ignored. This warning
>> message isn't related to dispatchable loads. Did you get any other error
>> messages?
>>
>> Your input for dispatachable load looks fine to me, except I think
>> Qmax (column 4) should probably be 0 to prevent your loads from acting
>> as 'reactive generators'.
>>
>> Kind regards,
>>
>> Samuel
>>
>> -Original Message-
>> From: bounce-120786253-76238...@list.cornell.edu
>> [mailto:bounce-120786253-76238...@list.cornell.edu] On Behalf Of
>> Muhammd Ali Mohseni
>> Sent: 14. september 2016 14:56
>> To: matpower-l@cornell.edu
>> Subject: benefit function of dispatchable load
>>
>> Dear all
>> According to the Matpower manual (6.4.2 Dispatchable Loads)  approach
>> to dispatchable  loads is to model them as negative real power
>> injections with associated negative costs. but when I run dcopf I got this
>> error!!!
>>
>> " MATPOWER Version 6.0b1, 01-Jun-2016 -- DC Optimal Power Flow The
>> interior-point-convex algorithm does not accept an initial point.
>> Ignoring X0."
>>
>> I set negative output, ranging from 0 to -1000 like this mpc.gen = [
>> .
>> 10   0   100 -1001.05100 1   0-1000  
>>   0 0   0   0   0   0   0   0   0   0
>>0;
>> ]
>> also I assume that bid function of load like this : aP+ b and benefit
>> function of load : 0.5aP^2+bP
>>
>> mpc.gencost = [
>> .
>> 20   0   3   -0.5*0.015  13.00;
>> ]
>>
>>
>
>

RE: benefit function of dispatchable load

[Samuel Perkin]

Dear Muhammd,

The coefficients of the cost function should be positive (see Figure 6-5 in the 
Matpower manual), and as dispatchable load is modelled as negative generation 
your cost value will be negative (i.e. a benefit). The OPF aims to minimize 
costs (or maximise negative costs) the model will now correctly aim to increase 
load as much as is economically feasible. 

In short, what you have now is correct!

Kind regards,

Samuel

-Original Message-
From: bounce-120786918-76238...@list.cornell.edu 
[mailto:bounce-120786918-76238...@list.cornell.edu] On Behalf Of Muhammd Ali 
Mohseni
Sent: 14. september 2016 17:19
To: MATPOWER discussion forum <matpowe...@list.cornell.edu>
Subject: Re: benefit function of dispatchable load

Dear Samuel,

I don't have any other problem. It is interesting when I change the benefit 
function coefficient to positive like this:
mpc.gencost = [
 .
 2   0   0   3   0.5*0.015  13.00;
 ]
there isn't any warning and DC Opf converge!!! now results is true???!!! hence 
the power associated with dispatchable Loads is negative but first part of cost 
function is positive(0.5*0.015*P^2>0)!! Is the final value of Opf correct???

With respect,

 Ali

On 9/14/16, Samuel Perkin <samu...@landsnet.is> wrote:
> Dear Muhammd,
>
> Your error message is not really an error, but is just a warning 
> statement that you shouldn't need to react to. It simply means that 
> the optimisation algorithm doesn't require an initial guess of the 
> solution, and therefore any initial guess is ignored. This warning 
> message isn't related to dispatchable loads. Did you get any other error 
> messages?
>
> Your input for dispatachable load looks fine to me, except I think 
> Qmax (column 4) should probably be 0 to prevent your loads from acting 
> as 'reactive generators'.
>
> Kind regards,
>
> Samuel
>
> -Original Message-
> From: bounce-120786253-76238...@list.cornell.edu
> [mailto:bounce-120786253-76238...@list.cornell.edu] On Behalf Of 
> Muhammd Ali Mohseni
> Sent: 14. september 2016 14:56
> To: matpower-l@cornell.edu
> Subject: benefit function of dispatchable load
>
> Dear all
> According to the Matpower manual (6.4.2 Dispatchable Loads)  approach 
> to dispatchable  loads is to model them as negative real power 
> injections with associated negative costs. but when I run dcopf I got this 
> error!!!
>
> " MATPOWER Version 6.0b1, 01-Jun-2016 -- DC Optimal Power Flow The 
> interior-point-convex algorithm does not accept an initial point.
> Ignoring X0."
>
> I set negative output, ranging from 0 to -1000 like this mpc.gen = [ .
> 1 0   0   100 -1001.05100 1   0-1000  
>   0 0   0   0   0   0   0   0   0   0 
>   0;
> ]
> also I assume that bid function of load like this : aP+ b and benefit 
> function of load : 0.5aP^2+bP
>
> mpc.gencost = [
> .
> 2 0   0   3   -0.5*0.015  13.00;
> ]
>
>

Re: benefit function of dispatchable load

[Muhammd Ali Mohseni]

Dear Samuel,

I don't have any other problem. It is interesting when I change the
benefit function coefficient to positive like this:
mpc.gencost = [
 .
 2   0   0   3   0.5*0.015  13.00;
 ]
there isn't any warning and DC Opf converge!!! now results is
true???!!! hence the power associated with dispatchable Loads is
negative but first part of cost function is
positive(0.5*0.015*P^2>0)!! Is the final value of Opf correct???

With respect,

 Ali

On 9/14/16, Samuel Perkin  wrote:
> Dear Muhammd,
>
> Your error message is not really an error, but is just a warning statement
> that you shouldn't need to react to. It simply means that the optimisation
> algorithm doesn't require an initial guess of the solution, and therefore
> any initial guess is ignored. This warning message isn't related to
> dispatchable loads. Did you get any other error messages?
>
> Your input for dispatachable load looks fine to me, except I think Qmax
> (column 4) should probably be 0 to prevent your loads from acting as
> 'reactive generators'.
>
> Kind regards,
>
> Samuel
>
> -Original Message-
> From: bounce-120786253-76238...@list.cornell.edu
> [mailto:bounce-120786253-76238...@list.cornell.edu] On Behalf Of Muhammd Ali
> Mohseni
> Sent: 14. september 2016 14:56
> To: matpower-l@cornell.edu
> Subject: benefit function of dispatchable load
>
> Dear all
> According to the Matpower manual (6.4.2 Dispatchable Loads)  approach to
> dispatchable  loads is to model them as negative real power injections with
> associated negative costs. but when I run dcopf I got this error!!!
>
> " MATPOWER Version 6.0b1, 01-Jun-2016 -- DC Optimal Power Flow The
> interior-point-convex algorithm does not accept an initial point.
> Ignoring X0."
>
> I set negative output, ranging from 0 to -1000 like this mpc.gen = [ .
> 1 0   0   100 -1001.05100 1   0-1000  
>   0 0   0   0   0   0   0   0   0   0 
>   0;
> ]
> also I assume that bid function of load like this : aP+ b and benefit
> function of load : 0.5aP^2+bP
>
> mpc.gencost = [
> .
> 2 0   0   3   -0.5*0.015  13.00;
> ]
>
>

RE: benefit function of dispatchable load

[Samuel Perkin]

Dear Muhammd,

Your error message is not really an error, but is just a warning statement that 
you shouldn't need to react to. It simply means that the optimisation algorithm 
doesn't require an initial guess of the solution, and therefore any initial 
guess is ignored. This warning message isn't related to dispatchable loads. Did 
you get any other error messages?

Your input for dispatachable load looks fine to me, except I think Qmax (column 
4) should probably be 0 to prevent your loads from acting as 'reactive 
generators'.

Kind regards,

Samuel

-Original Message-
From: bounce-120786253-76238...@list.cornell.edu 
[mailto:bounce-120786253-76238...@list.cornell.edu] On Behalf Of Muhammd Ali 
Mohseni
Sent: 14. september 2016 14:56
To: matpower-l@cornell.edu
Subject: benefit function of dispatchable load

Dear all
According to the Matpower manual (6.4.2 Dispatchable Loads)  approach to 
dispatchable  loads is to model them as negative real power injections with 
associated negative costs. but when I run dcopf I got this error!!!

" MATPOWER Version 6.0b1, 01-Jun-2016 -- DC Optimal Power Flow The 
interior-point-convex algorithm does not accept an initial point.
Ignoring X0."

I set negative output, ranging from 0 to -1000 like this mpc.gen = [ .
1   0   0   100 -1001.05100 1   0-1000  
  0 0   0   0   0   0   0   0   0   0   
0;
]
also I assume that bid function of load like this : aP+ b and benefit function 
of load : 0.5aP^2+bP

mpc.gencost = [
.
2   0   0   3   -0.5*0.015  13.00;
]