Load Shapes & Impact Profiles

Allow users to reference load shapes in their CET input data instead of impact profiles

This would eliminate the need to pre-calculate impact profiles for every possible combination of load shapes in advance. This would also allow dual baseline measures to use the correct impact profile for the early retirement phase and the above standard phase of their measure, instead of having to use one impact profile for both baselines as the CET works currently.

Remove NormUnit from load shape 8760 upload specification

Per request to align with the energy plus post processing script

Load Shape - Downloadable validation report

As a measure developer, I want to download the validation report in a portable format (CSV, Excel, text, or PDF), so that I can document issues, share errors with stakeholders, and track corrections across multiple upload attempts. Acceptance Criteria: A download button is available on the validation results page Supported formats include at least one of: CSV, Excel (.xlsx) The downloaded file includes all validation errors shown on screen The filename reflects the upload session The download is available immediately after validation completes and does not require re-running the validation

Support Baseline Tech Group and Tech Type in Load Shape Uploads

As a Measure Developer, I want CEDARS to recognise baseline technology information separately from measure technology information so that I can upload baseline load shapes using their correct Tech Group and Tech Type values. Background CEDARS currently only synchronise the measure case Tech Group and Tech Type from eTRM permutations. As a result, both measure and baseline load shapes are expected to use the measure case values during upload. For example: Measure: SWWH031-07 Measure Tech Type: HP_COP Baseline Tech Type: Stor_UEF Although the Tech ID correctly identifies the baseline technology, uploads currently fail unless the baseline load shape is generated using HP_COP. This behaviour has caused user confusion and ongoing troubleshooting effort.

Quarterly Avoided Cost graph is wonky while generating

Jake (PGE) noted that “When generating, the graph gets wonky (turns super long). Reloading - or changing the drop down - fixes it but maybe something that can be done later to make it look like that from the start (or force reload at the end). This should be low priority.”

Additional load shape derived impact profile validation

Request from Peter B’s team to add validation before impact profiles are generated/exported to confirm that each profile has exactly 8,760 rows no missing or duplicate hours hour_of_year values from 1 through 8,760 matching baseline and measure hour sequences input/output shapes that sum to 1.0 within an appropriate tolerance

Export Load Shape Library Table

As a user of the Load Shapes Library, I want to export the library table to CSV or Excel, so that I can more easily search, filter, and review the full list of load shapes outside the application. Description Add an export function that lets users download the load shape library table (currently containing thousands of entries) as a CSV or Excel file, since searching/filtering within the current table interface is difficult at this volume.

eTRM Sync – Skip Measure Packages with Missing Draft References

Some eTRM records may reference a draft_measure_package_id that does not have a corresponding draft Measure Package record. When this occurs: Skip the affected Measure Package and continue syncing valid Measure Packages. Show the skipped Measure Package on the eTRM Sync History page being released as part of @Improve eTRM Sync Resilience for Invalid Measure Package Data. Include the reason it was skipped, for example: draft_measure_package_id referenced in impact_profiles_latestpermutation was not found in impact_profiles_draftmeasurepackage.

Display UEC proportion values when selecting draft measure package

As a Measure Developer, I want to see proportion UEC values when selecting a Measure Package so that I can confirm I am uploading load shapes against the correct eTRM version. Description Multiple versions of the same Measure Package can appear in CEDARS following draft commits or discards in eTRM. The current Measure Package Values table does not display UEC proportion values, making it difficult to determine which version should be used. Selecting the wrong version may result in Impact Profiles and Avoided Cost Combinations being generated using outdated savings assumptions.

PA Measure Developer registration flow

Currently the development team can change user types to measure developers. This ticket would allow users to self-register. As a user involved in measure development, I want to register for a PA Measure Developer account and have my access request approved, so that I can upload load shapes and generate impact profiles

Load Shape User Guide

As a Measure Developer or Program Administrator, I want clear documentation explaining how to use the Load Shape Library tool, so that I can correctly prepare, upload, load shapes.

Export Hourly Data for a Specific Load Shape

It would be useful to add two separate functions to the load shape library for current/future users, particularly if the functionality of the load shape process allows for load shapes to be used between measure package (as is planned). First, it would be good to be able to export the library table itself as a csv or excel file. We have thousands of load shapes now, and searching/filtering in the current table can be challenging. Being able to export to excel and review there would be helpful. Second, being able to export the actual hourly data for a specific load shape in the library. This would be important for users who want to adopt load shapes between measure packages, or use them for other purposes, such as billing impact or other hourly analysis.

Allow measure packages to share load shapes

New review team file download

Request from Peter B’s team. New impact profile file for download that includes: a.  Impact Profile ID / E3MeaElecEndUseShape b.  PA / IOU c.  E3TargetSector d.  E3ClimateZone e.  avoided-cost version f.   avoided-cost combination ID / AC Elec Key g.  BldgType h.  BldgVint i.   BldgHVAC j.   Measure TechID k.  Baseline TechID l.   TechGroup and TechType, where applicable m. first-baseline vs second-baseline indicator, if applicable n.  source year/calendar used, e.g. 2009 non-leap Thursday-start o.  baseline load-shape ID used p.  measure load-shape ID used q.  baseline annual UEC used r.   measure annual UEC used s.  denominator/savings used in the impact formula t.   validation status/results u.  generation timestamp v. eTRM measure package ID w. eTRM permutation ID x. eTRM source snapshot/extract timestamp And also provides additional files containing: 1. baseline load-shape 8760 csv 2. measure load-shape 8760 csv Please give files descriptive names that include the Impact Profile ID or AC Elec Key.

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.

Use the modeled load shapes from one Measure Package for a non-modeled measure package

Background PG&E wants to reuse the Measure Case load shape from SWCR015 (Glassdoor-LED refrigerated cases) as both the Measure Case and Base Case load shape for SWCR018 (standard-efficiency → high-efficiency refrigerator/freezer replacement). Why SWCR015's ACC/load shape pair models "always-open case" (base) vs. "glass door installed" (measure) — a door-addition measure. SWCR018 is an efficiency-upgrade measure where both base and measure cases have doors; only unit efficiency changes. Applying SWCR015's shapes as-is to SWCR018 would incorrectly bake in the "doors vs. no doors" load shape difference, which doesn't reflect SWCR018's actual measure logic. Proposed workaround Set both the "Measure Tech ID" and "Standard Tech ID" fields in the eTRM permutation to the identical value (NE-Ref_Storage-VertDisplay-Glassdoor-LED-AirCond), differing only in UEC. This was tested locally and confirmed to work — CEDARS generates the impact profile fine as long as a permutation exists with matching Tech IDs on both sides and differing UEC values. Confirmed technically CEDARS has no structural issue generating an impact profile where Measure Tech ID = Standard Tech ID, provided the permutation is synced from eTRM with that configuration and UECs differ. Resulting ACC/Impact Profile ID: GroExAny | NE-Ref_Storage-VertDisplay-Glassdoor-LED-AirCond | NE-Ref_Storage-VertDisplay-Glassdoor-LED-AirCond Questions Precedent/policy question: Is reusing a measure-case load shape as both base and measure case an approach that should require sign-off from the broader load shapes group, given it's a workaround rather than a standard/intended use. Downstream clarity/misleading risk: When someone else (analyst, auditor, future CPUC reviewer) encounters an impact profile with identical Tech IDs on both the measure and base side, will they understand why? Scope/generalizability: Is this specific to SWCR018, or should we anticipate other measure packages with a similar "shape doesn't match base/measure logic" mismatch making the same request? If so, should we document a standard process/guidance for this pattern rather than a one-off answer?

Allow users to replace previously uploaded load shapes, impact profiles, and avoided costs

Currently load shapes are overwritten when re-uploaded. The generated impact profiles should be replaced. Background Measure Developers may need to update previously submitted load shapes due to: Corrections to uploaded data Changes to eTRM permutations or assumptions Re-running simulations with updated modeling inputs QC findings identified after initial upload Currently, CEDARS allows users to re-upload load shape files, which replaces the underlying load shape data. However, any previously generated Impact Profiles remain locked and cannot be recalculated. This can result in inconsistencies between the latest load shape data and the Impact Profiles and avoided cost combinations being used downstream. In addition, users currently receive an error when attempting to regenerate Impact Profiles that have already been generated: “Impact profiles have already been generated for the load shapes in this upload. Existing impact profiles cannot be recalculated.” This limitation prevents Measure Developers from correcting or updating previously submitted data. Proposed Enhancement Allow users to replace previously uploaded load shapes and regenerate associated Impact Profiles and avoided cost combinations so that CEDARS reflects the latest approved data. The enhancement should: Permit replacement of existing load shapes for a draft measure package. Allow existing Impact Profiles to be removed and regenerated from the updated load shapes. Provide transparency and safeguards to prevent unintended overwrites.

Duplicate Hour in Impact Profile Exports

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.

Deemed Measure Load Shape Repository

As a consultant to the CPUC, DNV is leading the effort to create a library of deemed measure load shapes on CEDARS. This is needed to enable the measure package developers and PAs to upload 2-3 load shapes for each deemed measure permutation in the eTRM for PY2028. These load shapes will be used to determine the quarterly avoided costs for each deemed measure claim as an input to the CET. It will be important to leverage the viewing/calculation capabilities that were developed for the impact profile library. We think there will also be a need to have a second sandbox library that would enable measure package developers to upload load shapes pertinent to a given measure package in preparation for their CET runs. This will need to be ready for use by measure package developers by mid-late spring 2026. This effort is intended to enable the replacement of the legacy DEER load shapes--many of which may no longer accurately represent the hourly distribution of savings--to improve the accuracy of the TSB and TRC of deemed measure.

Load Shapes – Version control between eTRM & Cedars

As a Measure Package Developer, I want CEDARS and eTRM to consistently handle revisions to draft measure packages that affect load shape identifying fields, so that I can update measure package permutations during review without creating duplicate, orphaned, or incorrectly associated load shapes. Description Measure developers regularly update draft measure packages throughout the ex-ante review process. These updates may change fields that are currently used by CEDARS to identify, match, and reuse load shapes. Example: A measure developer updates one or more load shape-related permutation fields in eTRM (e.g. TechID, Building Type, Building Vintage, HVAC Type, Climate Zone). The updated measure package may no longer match the previously uploaded load shape records. The current MVP supports re-uploading load shapes when a measure is re-modeled and assumes replacement based on load shape identifying fields. However, the long-term versioning strategy between eTRM measure package revisions and CEDARS load shape records has not been defined.

Load shape performance

I’m logging the issue that was discussed in the email thread initiated by Henry Liu (PG&E) on 6/30 regarding load shape processing performance. The load shapes uploaded on 6/30 remain in the processing queue as of 7/8. This delay is also impacting the ex-ante measure package review schedule and the CPUC/ex-ante team’s ability to determine next steps. The processing time was initially estimated at approximately 30 hours, with an additional 25 hours later identified. However, the actual processing duration has significantly exceeded those estimates. Notably, this issue pertains to only a single measure package, while several additional packages must be processed to support the DEER 2028 update. Please prioritize the resolution of this issue as soon as possible. Thank you.

Add email notification when load shape derived impact profile generation completes

It takes hours sometimes and we don’t want users to have to keep coming back to cedars to check if the job has completed.

Load Shape - Add a new user type for PA staff to upload load shapes

In the March 5 PCG meeting it came up that we may benefit from adding a new user type in CEDARS for users to upload load shapes but not have access to the other PA data reporting tools. New CEDARS user type: PA_MeasureDeveloper Granted community user privileges plus upload load shapes. Page permissions Load shape library: Public access View load shape graphs: Public access Add a new load shape: Any PA user Load shape upload history: Any PA user (Can only see own PA’s uploads) Permission Matrix Feature Access Load Shape Library Authenticated Users View Load Shape Details & Graphs Authenticated Users Add New Load Shape PA Admin, PA Ordinary, PA Measure Developer View Load Shape Upload History PA Admin, PA Ordinary, PA Measure Developer (Assigned PA only) Generate Load Shape Derived Impact Profiles PA Admin, PA Ordinary, PA Measure Developer Add Custom Impact Profiles PA Admin, PA Ordinary Edit Custom Impact Profiles PA Admin, PA Ordinary

SWWH014-09: Workaround for NR and NC permutations sharing the same load shape

As a Measure Developer, I want a supported way to use the same modeled load shape data across Existing and New vintages, so that CET can successfully run impact profiles where the modeled energy results are intentionally the same across vintages. Sophiya reported an issue with SWWH014-09 Heat Pump Water Heater, Residential. The CET run failed because an Impact Profile ID was not available for the New vintage for some measures. Following further clarification, the issue is not that the base and measure case use the same load shape. For SWWH014-09: Normal Replacement (NR / Existing) and New Construction (NC / New) have the same modeled savings and load shapes. eTRM permutations include both Existing (Ex) and New (New) vintages, resulting in expected Load Shape / Impact Profile IDs for both. The underlying modeling was only run for the NR / Existing building configuration. The NR energy results were then also used for the NC permutations in eTRM because the water heating measure was not expected to be affected by differences between the building vintages. As a result, the generated 8760 files only contain Load Shape IDs associated with the Existing (Ex) vintage. The corresponding New (New) vintage Load Shape IDs are therefore not included in the uploaded load shape files. CET fails when processing the NC permutations because the expected New vintage Impact Profiles cannot be generated from the uploaded data. Note we have a ticket in the backlog to @Allow measure packages to share load shapes

Load Shape – Impact profile generation can become Stuck After Errors

Impact Profile generation may enter an invalid state when processing encounters an error. Users may be unable to retry generation, with uploads remaining marked as processing, or incorrectly report that Impact Profiles have already been generated. Generation fails due to missing or invalid required data. Upload may remain in a “busy” state indefinitely. Users cannot re-run Impact Profile generation after correcting the issue. The system may incorrectly report that Impact Profiles already exist when the generation did not complete successfully. Expected Behavior Failed generation attempts should be rolled back cleanly. Users should be able to retry generation once the underlying issue is resolved. Existing Impact Profiles should be detected only when the generation completes successfully. Investigation Review generation error handling and rollback logic. Verify that the processing status is correctly updated after failures. Verify that the generated profile checks do not treat failed attempts as successful. Review queue handling when multiple uploads and generation requests are processed concurrently. For the production uploads already affected, we need to identify them, reset them, allow regeneration, and notify users.

Add PA-specific queues to PAs don't get in each others way in the load shape library

After this is done we can allow users to upload multiple files.

Increase max files size for load shape 8760 files

Currently set at 50MB.

Allow measure developers to upload more than one load shape 8760 file at a time

Requested by Henry L.