Skip to main content

Duplicate Hour in Impact Profile Exports

Bug Reports2026Impact Profiles & Load Shapes

Load shapes Impact Profile export 8759 unique hours instead of 8760 — one hour appears twice while another is missing entirely, resulting in the output still containing 8,760 rows but with a duplicated hour timestamp rather than a duplicated impact value.

Specifically, hour 1587 is missing and hour 1588 appears twice.

Likely related to daylight saving time handling.

Select the module in CEDARS
Select user type
Posted by cedars team ·
Status: In Testing14 comments

Log in to comment and vote

Comments14

  • Jake Richardson (PG&E)

    Team•

    Sep 15

    •

    Merged request

    •

    1 vote

    Hour of Impact

    When I put in a custom impact profile with hours 1-8760 without duplicates or missing hours, when I download the data from CEDARS after the fact the hours do not go from 1-8760 without duplicates, and instead have a duplicate and a missing hour. My guess is CEDARS has converted these to a date code after upload (for use in the calculations) and then re-converted to hour of impact for download without handling daylight savings time properly to prevent the duplicate / missing hour. Either way, the download from CEDARS should produce the exact same profile I uploaded and not require rework to put back into 1-8760 terms.

    In my example 1683 is missing and 1684 is a duplicate, which aligns with daylight savings on March 12, 2023. Strangely there is not any issues in the data for fall back on Nov 5, 2023.

    • cedars team

      Team•

      Sep 15

      Thanks for the additional information @Jake Richardson (PG&E). I’m merging this with a ticket for the same issue.

  • cedars team

    Team•

    Oct 22

    Hello @Jake Richardson (PG&E) , this work is now available on cedars-2 for testing.

    We have implemented changes to prevent new Impact Profiles from having duplicate/missing hours and have fixed hours for all existing (DEER, Custom, and Load Shape-derived) Impact Profiles.

    We would appreciate your verification that exports no longer contain duplicate/missing hours for:

    • existing DEER Impact Profiles

    • existing non-DEER Impact Profiles

    • new Custom Impact Profiles

    • new Load Shape Impact Profiles

    • Jake Richardson (PG&E)

      Team•

      Oct 22

      I would recommend looping someone else in for the new load shape impact profiles section - I haven’t done work in that module. Maybe @Henry.Liu can help us check that there are no more duplicates.

      • Henry.Liu

        •

        Oct 22

        Update:

        Looking at production and CEDARS-2, selecting the same load shape I was able to verify I see the error in production and CEDARS-2 does now contain the missing hour. I believe it is generating the output file correctly now.

        This was still pertaining to production. I will be checking CEDARS-2 with a follow up comment.

        Our most recent load shape upload on 9/28 still shows a missing (hour 1587) / duplicate hour (1588) from the load shape download in CEDARS which I downloaded today to double check

        • cedars team

          Team•

          Oct 22

          Thank you both!

          @Henry.Liu , we appreciate you checking these values for us.

          May we also request your verification on cedars-2 that:

          • The downloaded output is correct (no duplicate or missing hours) for existing DEER Impact Profiles

          • After generating Impact Profiles from a new Load Shape upload (via the Add Deemed Load Shapes page), the downloaded output is correct.

          • After creating new Custom Impact Profiles (through the Add Impact Profile page), the downloaded output is correct.

          • Henry.Liu

            •

            Oct 22

            @cedars team

            Deemed:

            Downloaded random impact profiles from the library and the hours are correct.

            I only had a limited amount of MP (listed as draft) and shapes (CSV files made with the parameters that pertain to those MP) available in CEDARS 2. I did upload shapes to MP SWCR007-06-draft that already had shapes and it allowed me to replace them. I then verified the hours in the output and it looks ok.

            If we want to do a start from scratch test, we will need to delete the shapes and start fresh.

            • cedars team

              Team•

              Oct 23

              @Henry.Liu , new hours created through the replacement process for Measure Packages with existing Load Shapes should be sufficient for the new Load Shape upload case, so this combined with your earlier output checks for existing Deemed Impact Profiles can conclude testing for Load Shape Impact Profiles.

              Thank you for your thorough testing on this!

              Additionally, would it be possible for you or @Jake Richardson (PG&E) to verify that the output contains correct hours for the non-Load Shape Impact Profiles?

              In particular, we are looking to confirm that:

              • Jake Richardson (PG&E)

                Team•

                Oct 23

                Yes, I was planning to test, was just taking care of some other priorities first. I checked existing DEER, existing custom profiles, and newly created custom profiles and they all appear to be good to me. The only variance I have now from what was input was existing e-10 rounding, which presumably is due to the number of decimals CEDARS has (and is acceptable)

                • cedars team

                  Team•

                  Oct 24

                  Thank you @Jake Richardson (PG&E) , and apologies for the extra follow ups.

                  We appreciate you verifying these cases as well as flagging the e-10 rounding difference.

  • jake.richardson@pge.com

    •

    Oct 23

    •

    Merged request

    •

    2 votes

    Impact Profile Library - DEER Impact Profiles have repeated hours

    DEER Impact Profiles have repeated hours (example: DEER HVAC Chillers repeats 1708 rather than following the hour_of_year format 1-8760. Presumably this was for daylight savings time but I had though we agreed with DNV that the profiles should follow the 1-8760 format like the ACC does.

    • cedars team

      Team•

      Jun 5, 2025

      It was mentioned in today’s demo meeting that this should not be a blocker for production release, but needs to be addressed.

    • rachel.murray@dnv.com

      •

      Oct 28, 2025

      Thanks for noting this, Jake. Sorry for not seeing this sooner. Do you know whether this also occurs in any other DEER impact profiles?

    • rachel.murray@dnv.com

      •

      Feb 9

      Since DNV didn’t produce the legacy DEER impact profiles, we never looked into the times used. Does someone need to modify the input files to comply with the CIP requirements? CEDARS will need them to generate the quarterly avoided cost combinations for PY2028 once the 2026 Electric Avoided Cost Calculator drops.