RE: MOST and ramp rates (was Re: Ramp AGC)

[Bainan Xia]

Thanks for the explanation! I’ve figured out which constraint is actually 
binding in my case.

Best,
Bainan

From: bounce-122375618-80810...@list.cornell.edu 
[mailto:bounce-122375618-80810...@list.cornell.edu] On Behalf Of Ray Zimmerman
Sent: Tuesday, March 13, 2018 7:11 AM
To: MATPOWER discussion forum 
Subject: MOST and ramp rates (was Re: Ramp AGC)

That’s basically true. To be more specific, RAMP_10 is used to determine the 
amount that a contingency dispatch can deviate from the corresponding base 
case, or to limit the reserve supplied by a unit when using fixed zonal reserve 
requirements. And RAMP_30 is used to determine how much a base case dispatch in 
a particular period can deviate from the base case dispatches in adjacent 
periods. And RAMP_AGC is not currently used for anything.

In MOST there are essentially two types of “reserve”: (1) contingency reserve, 
and (2) load-following ramp reserve. The contingency reserve, limited by 
RAMP_10, can be handled in one of two ways, either by specifying fixed zonal 
reserve requirements (in a deterministic problem) or by specifying explicit 
contingencies that must be covered (in a stochastic problem). This is discussed 
in Section 3.2 of the MOST User’s 
Manual
 and defined by equations (4.16)–(4.22). The load-following ramp reserve, 
limited by RAMP_30, is described in Section 3.5 and defined by equations 
(4.23)–(4.27).

Hope this helps,

Ray



On Mar 8, 2018, at 7:29 PM, Stephen Suffian 
> wrote:

I believe I've seen a similar question asked before. See this 
link
 which indicates that RAMP_10 is used for reserve calculations, RAMP_30 for 
dispatch scheduling, and RAMP_AGC Is not currently used by matpower according 
to this 
link.

On Thu, Mar 8, 2018 at 6:51 PM, Bainan Xia 
> wrote:
Hi,

I’m running a huge power system simulation using MOST recently. I’ve got some 
information on ‘ramp rate (% of the rated capacity per min)’ for each type of 
generator from a paper. I would like to add these constraints into my case. 
I’ve found a field in mpc.gen, called ‘RAMP_AGC’, which is ramp rate for load 
following/AGC (MW/min). I’m wondering whether that is the right place to fill. 
Or the following columns ‘RAMP_10’ and ‘RAMP_30’ are the right choices. I’m 
also a little confused on the difference between ‘RAMP_10’ and ‘RAMP_30’. What 
does the word ‘reserve’ mean in the documentation.

Thanks for the help ☺

Best,
Bainan

Re: AC OPF Convergence Issue

[Aakanksha Dubey]

Thank you for the guidance! The main issue was with the values I assumed
for Vm. My case is converging now.

sincerely,
Aakanksha *Dubey*


On Tue, Mar 13, 2018 at 9:52 AM, Ray Zimmerman  wrote:

> Making Vm = 1 means that the voltage is equal to its nominal value
> specified in the BASE_KV column of the bus matrix.
>
> Regarding relaxing line limits, first this is the one case where setting
> them to zero is interpreted as completely unconstrained. That is a
> reasonable thing to try in the process of trying to get a convergent case.
> The question of what sort of violations of the original limits are
> acceptable depends of course on the particular system and scenario you are
> trying to analyze/optimize.
>
> Ray
>
>
>
> On Mar 7, 2018, at 10:21 PM, Aakanksha Dubey  wrote:
>
> Thank You, Mirish and Shri for the guidance.
>
> Shri, the model is converging now. I understand that Vmax and Vmin limits
> should have been more relaxed. I have a couple of quick follow-up
> questions:
>
>- I had set the voltage magnitudes on per unit basis since the network
>voltages range from 69kV to 500kV. By now making Vm=1, wouldn't all
>voltages be set equal to 500kV?
>- Also, while relaxing line limits, what is the maximum acceptable
>percentage I should be increasing those values to?
>
>
> sincerely,
> Aakanksha *Dubey*
>
>
> On Wed, Mar 7, 2018 at 7:25 PM, Abhyankar, Shrirang G. 
> wrote:
>
>> The voltage magnitudes in MATPOWER data files, (bus(:,VM)), are given in
>> per unit and are generally close to 1.0. You have very low voltages for
>> many of the buses. Similarly, the VMAX and VMIN values are also in per unit
>> and generally they are 1.0 and 0.95, respectively. The following code
>> results in convergent OPF.
>>
>>
>>
>> mpc = loadcase(‘Transmission3’);
>>
>> define_constants;
>>
>> mpc.bus(:,VM) = 1.0;
>>
>> mpc.bus(:,VA) = 0.0;
>>
>> mpc.bus(:,VMAX) = 1.1;
>>
>> mpc.bus(:,VMIN) = 0.9;
>>
>> runopf(mpc);
>>
>>
>>
>> Also, note that when relaxing the limits set them to large values, not
>> zeros.
>>
>>
>>
>> Thanks,
>>
>> Shri
>>
>> Ph: (630) 252 0219 <(630)%20252-0219>
>>
>> www.mcs.anl.gov/~abhyshr
>>
>> *From: * on behalf of
>> Aakanksha Dubey 
>> *Reply-To: *MATPOWER discussion forum 
>> *Date: *Wednesday, March 7, 2018 at 4:23 PM
>> *To: *"MATPOWER-L@cornell.edu" 
>> *Subject: *AC OPF Convergence Issue
>>
>>
>>
>> Hi All,
>>
>>
>>
>> I am developing a model for transmission grid at Lehigh Valley. My model
>> seems to converge for DC OPF but is unable to converge for AC. I have tried
>> the following to make it converge but nothing works:
>>
>>
>>
>> · Relaxed Vmax, Vmix, BR_R, Rate_A,Rate_B,Rate_C
>>
>> · Checked the tap ratio, Qmax, Qmin,Pmax,Pmin
>>
>> · Tried to change the OPF options by using opf.init_from_mpc
>>
>> · Using mpoption I tried to change the solver to IPOPT& MINOPF
>> but MATLAB keeps giving me an error that I need to download it. After
>> downloading and connecting the path to MATLAB it gives me the same error.
>>
>> I have attached my casefiles to this email. Please let me know if anyone
>> can find an issue with the model.
>>
>>
>>
>> Thank you!
>>
>>
>>
>> Aakanksha
>>
>>
>>
>
>
>

MOST and ramp rates (was Re: Ramp AGC)

[Ray Zimmerman]

That’s basically true. To be more specific, RAMP_10 is used to determine the 
amount that a contingency dispatch can deviate from the corresponding base 
case, or to limit the reserve supplied by a unit when using fixed zonal reserve 
requirements. And RAMP_30 is used to determine how much a base case dispatch in 
a particular period can deviate from the base case dispatches in adjacent 
periods. And RAMP_AGC is not currently used for anything.

In MOST there are essentially two types of “reserve”: (1) contingency reserve, 
and (2) load-following ramp reserve. The contingency reserve, limited by 
RAMP_10, can be handled in one of two ways, either by specifying fixed zonal 
reserve requirements (in a deterministic problem) or by specifying explicit 
contingencies that must be covered (in a stochastic problem). This is discussed 
in Section 3.2 of the MOST User’s Manual 
 and defined by 
equations (4.16)–(4.22). The load-following ramp reserve, limited by RAMP_30, 
is described in Section 3.5 and defined by equations (4.23)–(4.27).

Hope this helps,

Ray


> On Mar 8, 2018, at 7:29 PM, Stephen Suffian  wrote:
> 
> I believe I've seen a similar question asked before. See this link 
>  which 
> indicates that RAMP_10 is used for reserve calculations, RAMP_30 for dispatch 
> scheduling, and RAMP_AGC Is not currently used by matpower according to this 
> link .
> 
> On Thu, Mar 8, 2018 at 6:51 PM, Bainan Xia  > wrote:
> Hi,
> 
>  
> 
> I’m running a huge power system simulation using MOST recently. I’ve got some 
> information on ‘ramp rate (% of the rated capacity per min)’ for each type of 
> generator from a paper. I would like to add these constraints into my case. 
> I’ve found a field in mpc.gen, called ‘RAMP_AGC’, which is ramp rate for load 
> following/AGC (MW/min). I’m wondering whether that is the right place to 
> fill. Or the following columns ‘RAMP_10’ and ‘RAMP_30’ are the right choices. 
> I’m also a little confused on the difference between ‘RAMP_10’ and ‘RAMP_30’. 
> What does the word ‘reserve’ mean in the documentation.
> 
>  
> 
> Thanks for the help J
> 
>  
> 
> Best,
> 
> Bainan
> 
> 

Re: AC OPF Convergence Issue

[Ray Zimmerman]

Making Vm = 1 means that the voltage is equal to its nominal value specified in 
the BASE_KV column of the bus matrix.

Regarding relaxing line limits, first this is the one case where setting them 
to zero is interpreted as completely unconstrained. That is a reasonable thing 
to try in the process of trying to get a convergent case. The question of what 
sort of violations of the original limits are acceptable depends of course on 
the particular system and scenario you are trying to analyze/optimize.

Ray



> On Mar 7, 2018, at 10:21 PM, Aakanksha Dubey  wrote:
> 
> Thank You, Mirish and Shri for the guidance. 
> 
> Shri, the model is converging now. I understand that Vmax and Vmin limits 
> should have been more relaxed. I have a couple of quick follow-up questions:
> I had set the voltage magnitudes on per unit basis since the network voltages 
> range from 69kV to 500kV. By now making Vm=1, wouldn't all voltages be set 
> equal to 500kV? 
> Also, while relaxing line limits, what is the maximum acceptable percentage I 
> should be increasing those values to?
> 
> sincerely, 
> Aakanksha Dubey
> 
> 
> On Wed, Mar 7, 2018 at 7:25 PM, Abhyankar, Shrirang G.  > wrote:
> The voltage magnitudes in MATPOWER data files, (bus(:,VM)), are given in per 
> unit and are generally close to 1.0. You have very low voltages for many of 
> the buses. Similarly, the VMAX and VMIN values are also in per unit and 
> generally they are 1.0 and 0.95, respectively. The following code results in 
> convergent OPF.
> 
>  
> 
> mpc = loadcase(‘Transmission3’);
> 
> define_constants;
> 
> mpc.bus(:,VM) = 1.0;
> 
> mpc.bus(:,VA) = 0.0;
> 
> mpc.bus(:,VMAX) = 1.1;
> 
> mpc.bus(:,VMIN) = 0.9;
> 
> runopf(mpc);
> 
>  
> 
> Also, note that when relaxing the limits set them to large values, not zeros.
> 
>  
> 
> Thanks,
> 
> Shri
> 
> Ph: (630) 252 0219 
> www.mcs.anl.gov/~abhyshr 
> From:  > on behalf of Aakanksha 
> Dubey >
> Reply-To: MATPOWER discussion forum  >
> Date: Wednesday, March 7, 2018 at 4:23 PM
> To: "MATPOWER-L@cornell.edu " 
> >
> Subject: AC OPF Convergence Issue
> 
>  
> 
> Hi All, 
> 
>  
> 
> I am developing a model for transmission grid at Lehigh Valley. My model 
> seems to converge for DC OPF but is unable to converge for AC. I have tried 
> the following to make it converge but nothing works: 
> 
>  
> 
> · Relaxed Vmax, Vmix, BR_R, Rate_A,Rate_B,Rate_C
> 
> · Checked the tap ratio, Qmax, Qmin,Pmax,Pmin
> 
> · Tried to change the OPF options by using opf.init_from_mpc
> 
> · Using mpoption I tried to change the solver to IPOPT& MINOPF but 
> MATLAB keeps giving me an error that I need to download it. After downloading 
> and connecting the path to MATLAB it gives me the same error. 
> 
> I have attached my casefiles to this email. Please let me know if anyone can 
> find an issue with the model. 
> 
>  
> 
> Thank you!
> 
>  
> 
> Aakanksha
> 
>  
> 
>