Centre for Energy and Transportation Statistics
Energy Section

Transmission pipelines are establishments primarily engaged in the pipeline transportation of natural gas from gas fields or processing plants to local distribution systems.

Confidentiality

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Statistics Canada will use the information from this survey for statistical purposes.

Table of contents

• A – General Information: Purpose of survey, Data-sharing agreements and Data linkage
• B – Reporting Instructions
• C – Supply of Natural Gas Unit of Measure
• D – Imports
• E – Receipts from Domestic Sources
• F – Average Heating Value in Gigajoules/Thousand Cubic Meters
• G – Exports (Specify Port of Exit)
• H – Domestic Deliveries
• I – Report Amounts of Gas Delivered to Consumers
• J – Liquefied Natural Gas (LNG) Marine Terminals
• K – Consumed Own Fuel
• L – Line Pack Fluctuation
• M – Metering Differences, Line Loss, Other Unaccounted Adjustments
• N – In-transit Shipments of Natural Gas
• O – Ex-transit Shipments
• P – Thousands of Cubic Metre Kilometres (10³m³km)

A – General information

Purpose of survey

The purpose of this survey is to obtain information on the supply of, and demand for, energy in Canada. This information serves as an important indicator of Canadian economic performance, and is used by all levels of government in establishing informed policies in the energy area. In the case of public utilities, it is used by governmental agencies to fulfill their regulatory responsibilities. The private sector also uses this information in the corporate decision-making process. Your information may also be used by Statistics Canada for other statistical and research purposes.

Data-sharing agreements

To reduce respondent burden, Statistics Canada has entered into data-sharing agreements with provincial and territorial statistical agencies and other government organizations, which have agreed to keep the data confidential and use them only for statistical purposes. Statistics Canada will only share data from this survey with those organizations that have demonstrated a requirement to use the data.

Section 11 of the Statistics Act provides for the sharing of information with provincial and territorial statistical agencies that meet certain conditions. These agencies must have the legislative authority to collect the same information, on a mandatory basis, and the legislation must provide substantially the same provisions for confidentiality and penalties for disclosure of confidential information as the Statistics Act. Because these agencies have the legal authority to compel businesses to provide the same information, consent is not requested and businesses may not object to the sharing of the data.

For this survey, there are Section 11 agreements with the provincial and territorial statistical agencies of Newfoundland and Labrador, Nova Scotia, New Brunswick, Quebec, Ontario, Manitoba, Saskatchewan, Alberta, British Columbia, and the Yukon.

The shared data will be limited to information pertaining to business establishments located within the jurisdiction of the respective province or territory.

Section 12 of the Statistics Act provides for the sharing of information with federal, provincial or territorial government organizations. Under Section 12, you may refuse to share your information with any of these organizations by writing a letter of objection to the Chief Statistician and returning it with the completed questionnaire. Please specify the organizations with which you do not want to share your data.

For this survey, there are Section 12 agreements with the statistical agencies of Prince Edward Island, the Northwest Territories and Nunavut as well as with the provincial and territorial government ministries responsible for the energy sector, the Canada Energy Regulator, Natural Resources Canada, and Environment and Climate Change Canada.

For agreements with provincial and territorial government organizations, the shared data will be limited to information pertaining to business establishments located within the jurisdiction of the respective province or territory.

Data linkage

To enhance the data from this survey, Statistics Canada may combine it with information from other surveys or from administrative sources.

B – Reporting Instructions

Please report information for a specific reference month 2025.

Please complete all sections as applicable.

If the information requested is unknown, please provide your best estimate.

Value (cost to customer): reported dollar values should exclude all taxes. Further, rebates paid to the customer should be deducted to arrive at "value".

This guide is designed to assist you as you complete the Monthly Natural Gas Transmission Survey. If you need more information, please call 1-877-604-7828.

Supply

C – Supply of Natural Gas Unit of Measure

Amounts: report amounts (1000m³ or Gigajoules) of natural gas received and delivered during the month under review.

D – Imports

Report total amount of natural gas carried into Canada, by port of entry.

Inclusion: amounts of gas moving in transit (E.g.: from the U.S., through Canada, and back into the U.S.)

Exclusion: Receipts from Liquefied Natural Gas (LNG) marine terminals

E – Receipts from Domestic Sources

Report volumes of gas received from sources such as:

Fields

Report amounts of gas received from fields connected directly to your company's transmission system. Field flared and waste and re-injection should be deducted from this amount.

Field plants  

Report amounts of gas received at the processing or re-processing plant gate after the deduction of shrinkage, plant uses and losses.

Exclusions:

• Natural Gas Liquids (NGL) fractionation plants
• Mainline straddle plants;

 Gas gathering systems

Report amounts of gas received from gas gathering systems connected directly to your company's transmission system.

Natural Gas Liquids (NGL) fractionation plants and mainline straddle plants

Exclusion:

• Field gas plants

Other transmission pipelines

Report amounts of gas received from other transmission pipelines (NAICS 486210) connected directly to your company's transmission system.

Transmission pipelines are establishments primarily engaged in the pipeline transportation of natural gas from gas fields or processing plants to local distribution systems.

Storage facilities

Report amounts of gas received from storage facilities (NAICS 493190) connected directly to your company's transmission system.

Storage facilities include natural gas storage caverns and liquefied natural gas storage, but exclude establishments primarily engaged liquefaction and regasification of natural gas for purposes of transport (NAICS 488990).

Distributors (utility distribution systems)

Report amounts of gas received from gas distributors (NAICS 221210) connected directly to your company's transmission system.

Gas distributors are establishments primarily engaged in the distribution of natural or synthetic gas to the ultimate consumers through a system of mains.

Liquefied Natural Gas (LNG) marine terminals

Report amounts of gas received from LNG marine terminals (NAICS 488990) connected directly to your company's transmission system.

LNG marine terminals are establishments primarily engaged with the liquefaction and regasification of natural gas for purposes of transport.

F – Average Heating Value in Gigajoules/ Thousand Cubic Meters

Report average heat content of your natural gas receipts for the reported reference month.

Disposition

G – Exports, Specify Port of Exit

Report total amount of natural gas this transmission pipeline physically exported from Canada to the United States, by port of exit.

Inclusion: amounts of gas moving (E.g.: from Canada, through the U.S., and back into Canada)

Exclusion: Deliveries to Liquefied Natural Gas (LNG) marine terminals

H – Domestic Deliveries

Report amount of natural gas delivered to facilities and pipelines such as:

Natural Gas Liquids (NGL) fractionation plants and mainline straddle plants

Exclusion:

• Field gas plants
• Other transmission pipelines

Report amounts of gas delivered to other transmission pipelines (NAICS 486210) connected directly to your company's transmission system.

Transmission pipelines are establishments primarily engaged in the pipeline transportation of natural gas from gas fields or processing plants to local distribution systems.

Storage facilities

Report amounts of gas delivered to storage facilities (NAICS 493190) connected directly to your company's transmission system.

Storage facilities include natural gas storage caverns and liquefied natural gas storage but exclude establishments primarily engaged liquefaction and regasification of natural gas for purposes of transport (NAICS 488990).

Distributors (utility distribution systems)

Report amounts of gas delivered to gas distributors (NAICS 221210) connected directly to your company's transmission system.

Gas distributors are establishments primarily engaged in the distribution of natural or synthetic gas to the ultimate consumers through a system of mains.

I – Report Amounts of Gas Delivered to Consumers and report the number of customers

Industrial power generation plants

Report gas delivered to electric power generation plants (NAICS 2211) connected directly to your company's transmission system.

This industry comprises establishments primarily engaged in the generation of bulk electric power by natural gas.

Other industrial

Deliveries to Other Industrial Consumers

Report gas delivered to industrial establishments other than power generation plants.

Inclusions:

• Agriculture and forestry
• Mining, quarrying, and oil and gas extraction
• Construction
• Manufacturing

Exclusions:

• Electric power generation
• Wholesale and retail trade
• Transportation and warehousing
• Other commercial buildings (e.g., public institutions)
• Natural gas transmission pipelines
• Natural gas storage facilities
• Natural gas distributors

Commercial and institutional

Report gas delivered to commercial and institutional establishments.

Inclusions:

• Wholesale and retail trade
• Transportation and warehousing
• Other commercial buildings (e.g., public institutions)

Value (cost to customer): dollar values exclude provincial taxes (if applicable), goods and services tax (GST) and harmonized sales tax (HST). Further, rebates paid to the customer should be deducted in order to arrive at "value".

J – Liquefied Natural Gas (LNG) Marine Terminals

Report amounts of gas delivered to LNG marine terminals (NAICS 488990) connected directly to your company's transmission system.

LNG marine terminals are establishments primarily engaged liquefaction and regasification of natural gas for purposes of transport.

K – Consumed Own Fuel

Report amount of gas consumed to fuel this transmission system.

L – Line Pack Fluctuation

Report the change in line pack between the first and last day of the reference month.

M – Metering Differences, Line Loss, Other Unaccounted Adjustments

Report the difference between the total supply and total disposition. This difference includes leakage or other losses, discrepancies due to metering inaccuracies and other variants, particularly billing lag.

N – In-transit Shipments of Natural Gas

Report total amount of natural gas received into Canada with the intention of exporting it back to the United States. (Re-Export)

O – Ex-transit Shipments

Report total amount of natural gas delivered to the United States with the intention of importing it back to Canada. (Re-Import)

P – Thousands of Cubic Metre Kilometres (10³m³km)

Please report the volume of natural gas transmitted (in 10³m³) multiplied by the distance (in km) each shipment has travelled.

Example:

• Step 1) 2 000 cubic metres transported over 5 km is equal to 10 000 cubic metre km.
• Step 2) To report in 10³m³km, divide 10 000 cubic metre km by 1 000, which equals 10 cubic metre km.

Thank you for your participation.

Wildfires

• On May 13th, the Government of Manitoba announced that, with wildfire conditions continuing to escalate, it had declared a state of local emergency for Nopiming, Wallace Lake, South Atikaki and Manigotagan River provincial parks and issued an evacuation order for Nopiming Provincial Park. The Government said these parks would remain closed to the public until conditions improve. On May 14th, the Government announced it was closing Whiteshell Provincial Park until conditions improve. On May 28th, the Government of Manitoba declared a provincewide state of emergency, effective for 30 days. The Government also said that in addition, the city of Flin Flon and the First Nations of Pimicikimak and Mathias Colomb had issued mandatory evacuation orders.
• On May 29th, the Government of Saskatchewan declared a provincial State of Emergency due to the wildfires affecting communities across Saskatchewan. The Government said the state of emergency would be in effect for 30 days.

Resources

• Texas-based Sunoco LP and Parkland Corporation of Calgary announced they had entered into a definitive agreement whereby Sunoco will acquire all outstanding shares of Parkland in a cash and equity transaction valued at approximately USD $9.1 billion, including assumed debt. The companies said the transaction is expected to close in the second half of 2025, subject to the satisfaction of closing conditions, including approval by Parkland's shareholders and customary regulatory and stock exchange listing approvals.
• France-based TotalEnergies SE announced it had signed a Sales and Purchase Agreement with Ksi Lisims LNG for the purchase of 2 Mtpa (million tonnes per annum) of LNG for 20 years from the future liquefaction plant located on the northwest coast British Columbia. The company said that, in parallel, it acquired a 5% stake in Texas-based Western LNG LLC, the developer, shareholder, and future operator of the Ksi Lisims LNG project.
• Calgary-based Strathcona Resources Ltd. announced it had entered into definitive agreements to sell substantially all of its Montney assets for approximately $2.84 billion to ARC Resources Ltd. and Tourmaline Oil Corp.
• Calgary-based Vermilion Energy Inc. announced it had entered into a definitive agreement for the sale of its Saskatchewan and Manitoba assets for cash proceeds of $415 million. Vermillion said the transaction is anticipated to close in the third quarter of 2025, subject to receipt of regulatory approvals and the satisfaction of other customary closing conditions.
• Vancouver-based Pan American Silver Corp. and MAG Silver Corp. announced they had entered into a definitive agreement whereby Pan American will acquire all of the issued and outstanding common shares of MAG for total consideration of approximately USD $2.1 billion. The companies said the transaction is expected to close in the second half of 2025, subject to the satisfaction of customary closing conditions, including shareholder approval, clearance under Mexican anti-trust laws, and approval of the listing of the Pan American common shares to be issued under the transaction on both the Toronto Stock Exchange and the New York Stock Exchange.
• The Government of Ontario announced it had approved Ontario Power Generation's (OPG) plan to begin construction on the first of four small modular reactors (SMRs) at the Darlington nuclear site. The Government said the SMR will produce enough electricity to power the equivalent of 300,000 homes.

Canada's internal trade

• The Government of Manitoba announced it had signed an agreement with the Ontario government to boost the flow of goods, services, investment and workers in both provinces. The Government said the premiers signed a memorandum of understanding to signal the provinces' intention to work together to knock down interprovincial trade barriers including direct-to-consumer alcohol sales and improved labour mobility between Manitoba and Ontario.
• The Government of Nova Scotia announced it was removing more interprovincial trade barriers by introducing legislative changes to enhance the new Traffic Safety Act in the fall that would allow more types of commercial trucks and other passenger vehicles to enter and operate in the province, supporting the movement of goods and services across the country. The Government also said it plans to amend the Nova Scotia Building Code Regulations to allow factory-built (modular) buildings that meet the National Building Code to be installed in the province without having to meet additional Nova Scotia-specific standards.
• The Government of Quebec announced it had tabled a Bill to promote trade in products and the mobility of labour from other provinces and territories of Canada, in order to stimulate interprovincial trade by improving the free movement of goods and skilled workers.

Other news

• The Government of Canada announced that it was moving forward with the proposal to deliver tax relief for Canadians by reducing the lowest marginal personal income tax rate from 15% to 14%, effective July 1, 2025.
• The Government of Alberta announced it was freezing the industrial carbon price at the current rate of $95 per tonne of emissions to keep industry competitive and defend jobs in response to the uncertainty caused by United States' tariffs.
• On May 15th, the Government of Ontario released its 2025 Budget, which included support for workers and businesses, the creation of a new Critical Minerals Processing Fund, and investments in infrastructure, skills training, housing, health, and education. The Government forecasts a $14.6 billion deficit in 2025-26 and real GDP growth of 0.8% in 2025 and 1.0% in 2026.
• The Government of Ontario announced its plan to permanently cut the gasoline and fuel tax rates, keeping the provincial rates of tax at nine cents per litre, which the Government said would save households, on average, about $115 per year going forward. The Government also said it was proposing to remove tolls from the provincially owned Highway 407 East, which the Government said is expected to save daily commuters an estimated $7,200 annually.
• Quebec's minimum wage increased from $15.75 to $16.10 per hour on May 1st.
• The Government of the Yukon announced that the 2025 rent index would be set at 2.0%.
• Toronto-based Canadian Tire Corporation announced it had entered into a definitive agreement to become the home of Canadian brands and other intellectual property of the Hudson's Bay Company (HBC), including the HBC Stripes and various company names, logos, designs, coat of arms and brand trademarks, for $30 million. Canadian Tire said the transaction is expected to close later this summer, subject to court approval and other customary terms and conditions.
• In a court filing with the Superior Court of Ontario, the Hudson's Bay Company (HBC) said that by June 1, 2025, it will have terminated approximately 8,347 or approximately 89% of its employees. HBC said the remaining 1,017 include Distribution Centre employees, and that the Distribution Centres are expected to close June 15, 2025.
• Toronto-based TD Bank Group announced it had initiated a new restructuring program in the second quarter of 2025 to reduce its cost base, including savings from an approximate 2% workforce reduction.
• Vancouver-based Telus Corporation announced it was investing more than $70 billion over the next five years to expand and enhance its network infrastructure and operations, including bringing TELUS PureFibre connectivity to homes and businesses across British Columbia, Alberta, Quebec, and Ontario; deploying enhancements to its 5G and LTE services; and launching two Sovereign AI Factories in Kamloops and Rimouski.
• Montreal-based BCE Inc. announced Bell AI Fabric, an investment that will create a national network starting with a data centre supercluster in British Columbia that will aim to provide upwards of 500 MW of hydro-electric powered AI compute capacity across six facilities. Bell said the first facility would come online in June 2025, in Kamloops BC with a second facility opening in Merritt, BC, by the end of 2025.
• Waterloo Ontario-based Definity Financial Corporation announced today that it had entered into a definitive agreement with the Travelers Companies, Inc. of New York to acquire Canadian operations of Travelers for cash consideration of approximately $3.3 billion. Definity said the transaction is expected to close in the first quarter of 2026, subject to customary regulatory approvals.

United States and other international news

• The U.S. Federal Open Market Committee (FOMC) maintained the target range for the federal funds rate at 4.25% to 4.50%. The last change in the target range was a 25 basis points cut in December 2024. The Committee also said that it would continue reducing its holdings of Treasury securities and agency debt and agency mortgage-backed securities.
• The Bank of Japan announced it will encourage the uncollateralized overnight call rate to remain at around 0.50%. The last change in the uncollateralized overnight call rate was a 25 basis points increase in January 2025.
• The Bank of England's Monetary Policy Committee (MPC) voted to reduce the Bank Rate by 25 basis points to 4.25%. The last change in the Bank Rate was a 25 basis points cut in February 2025.
• The Monetary Policy and Financial Stability Committee of Norway's Norges Bank left the policy rate unchanged at 4.5%. The last change in the policy rate was a 25 basis points increase in December 2023.
• The Executive Board of Sweden's Riksbank left the repo rate unchanged at 2.25%. The last change in the repo rate was a 25 basis points reduction in January 2025.
• The Reserve Bank of Australia (RBA) lowered the cash rate target by 25 basis points to 3.85%. The last change in the cash rate target was a 25 basis points cut in February 2025.
• The Reserve Bank of New Zealand (RBNZ) lowered the Official Cash Rate (OCR), its main policy rate, by 25 basis points to 3.25%. The last change in the OCR was a 25 basis points cut in April 2025.
• U.S. President Donald J. Trump and United Kingdom Prime Minister Sir Keir Starmer announced on May 8th a trade deal – the U.S.-UK Economic Prosperity Deal (EPD) – to remove barriers to make it easier for American and British businesses to operate, invest and trade in both countries.
• The White House announced on May 12th that U.S. President Donald J. Trump had reached an agreement with China whereby the United States would remove the additional tariffs it imposed on China on April 8 and April 9, 2025, but would retain all duties imposed on China prior to April 2, 2025. The White House said China would remove the retaliatory tariffs it announced since April 4, 2025, and would also suspend or remove the non-tariff countermeasures taken against the United States since April 2, 2025. President Trump said the suspension of tariffs would be for a period of 90 days.
• Moody's Ratings announced it had downgraded the Government of United States of America's long-term issuer and senior unsecured ratings to Aa1 from Aaa and changed the outlook from stable to negative. Moody's said the downgrade reflects the increase over more than a decade in government debt and interest payment ratios to levels that are significantly higher than similarly rated sovereigns.
• Virginia-based Boeing and Qatar Airways announced the carrier would purchase up to 210 widebody jets, including 130 787 Dreamliners, 30 777-9s, and options for an additional 50 787 and 777X airplanes. Boeing said the order would support nearly 400,000 jobs in the U.S.
• The eight OPEC+ countries - Saudi Arabia, Russia, Iraq, UAE, Kuwait, Kazakhstan, Algeria, and Oman - which previously announced additional voluntary adjustments in April and November 2023, announced they would implement a production adjustment of 411 thousand barrels per day, equivalent to three monthly increments, in June 2025.
• Japan-based Nissan Motor Co., Ltd announced a recovery plan with a cost reduction target of ¥250 billion. Nissan said it will consolidate its vehicle production plants from 17 to 10 by fiscal year 2027 and reduce its workforce by a total of 20,000 employees between fiscal years 2024 and 2027, which includes a previously announced reduction of 9,000.

Financial market news

• West Texas Intermediate crude oil closed at USD $60.79 per barrel on May 30th, up from a closing value of USD $58.21 at the end of April. Western Canadian Select crude oil traded in the USD $44 to $52 per barrel range throughout May. The Canadian dollar closed at 72.68 cents U.S. on May 30th, up from 72.40 cents U.S. at the end of April. The S&P/TSX composite index closed at 26,175.05 on May 30th, up from 24,841.68 at the end of April.

Centre for Energy and Transportation Statistics
Energy Section

Gas distributors are establishments primarily engaged in the distribution of natural or synthetic gas to the ultimate consumers through a system of mains.

Confidentiality

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Statistics Canada will use the information from this survey for statistical purposes.

Table of contents

• A – General Information: Purpose of survey, Data-sharing agreements and Data linkage
• B – Reporting Instructions
• C – Supply of Natural Gas Unit of Measure
• D – Receipts from Transmission Pipelines
• E – Receipts from Storage Facilities
• F – Receipts from Other Gas Distributors
• G – Total Supply of Natural Gas
• H – Average Heating Value in Gigajoules/Thousand Cubic Meters
• I – Deliveries to System Gas Consumers
• J – Deliveries to Consumers Enrolled with a Third Party Marketer
• K – Deliveries to Consumers who have Purchased Directly from Suppliers
• L – Deliveries to Power Generation Plants
• M – Deliveries to Other Industrial Consumers
• N – Deliveries to Commercial and Institutional Consumers
• O – Deliveries to Residential Consumers
• P – Deliveries to Transmission Pipelines
• Q – Deliveries to Storage Facilities
• R – Deliveries to Other Gas Distributors
• S – Own Use
• T – Line Pack Fluctuation
• U – Metering Differences, Line Loss, Other Unaccounted for and Cyclical Billing Adjustments
• V – Average Heating Value in Gigajoules/Thousand Cubic Meters
• W – Total Disposition

A - General information

Purpose of survey

The purpose of this survey is to obtain information on the supply of, and demand for, energy in Canada. This information serves as an important indicator of Canadian economic performance and is used by all levels of government in establishing informed policies in the energy area. In the case of public utilities, it is used by governmental agencies to fulfill their regulatory responsibilities. The private sector also uses this information in the corporate decision-making process. Your information may also be used by Statistics Canada for other statistical and research purposes.

Data-sharing agreements

To reduce respondent burden, Statistics Canada has entered into data-sharing agreements with provincial and territorial statistical agencies and other government organizations, which have agreed to keep the data confidential and use them only for statistical purposes. Statistics Canada will only share data from this survey with those organizations that have demonstrated a requirement to use the data.

Section 11 of the Statistics Act provides for the sharing of information with provincial and territorial statistical agencies that meet certain conditions. These agencies must have the legislative authority to collect the same information, on a mandatory basis, and the legislation must provide substantially the same provisions for confidentiality and penalties for disclosure of confidential information as the Statistics Act. Because these agencies have the legal authority to compel businesses to provide the same information, consent is not requested, and businesses may not object to the sharing of the data.

For this survey, there are Section 11 agreements with the provincial and territorial statistical agencies of Newfoundland and Labrador, Nova Scotia, New Brunswick, Quebec, Ontario, Manitoba, Saskatchewan, Alberta, British Columbia, and the Yukon.

The shared data will be limited to information pertaining to business establishments located within the jurisdiction of the respective province or territory.

Section 12 of the Statistics Act provides for the sharing of information with federal, provincial or territorial government organizations. Under Section 12, you may refuse to share your information with any of these organizations by writing a letter of objection to the Chief Statistician and returning it with the completed questionnaire. Please specify the organizations with which you do not want to share your data.

For this survey, there are Section 12 agreements with the statistical agencies of Prince Edward Island, the Northwest Territories and Nunavut as well as with the provincial and territorial government ministries responsible for the energy sector, the Canada Energy Regulator, Natural Resources Canada, and Environment and Climate Change Canada.

For agreements with provincial and territorial government organizations, the shared data will be limited to information pertaining to business establishments located within the jurisdiction of the respective province or territory.

Data linkage

To enhance the data from this survey, Statistics Canada may combine it with information from other surveys or from administrative sources.

B - Reporting Instructions

Please report information for a specific reference month 2025.

Please complete all sections as applicable.

If the information requested is unknown, please provide your best estimate.

Amounts: Report amounts in Gigajoules (GJs) of natural gas received and delivered during the month under review.

Value (cost to customer): The reported dollar value should exclude all taxes. Further, rebates paid to the customer should be deducted to arrive at "value".

This guide is designed to assist you as you complete the Monthly Natural Gas Distribution Survey. If you need more information, please call 1-877-604-7828.

C - Supply of Natural Gas Unit of Measure

Amounts: report amounts (1000m3 or Gigajoules) of natural gas received and delivered during the month under review.

D - Receipts from Transmission Pipelines

Report volumes of gas received from transmission pipelines (NAICS 486210) connected directly to your company's distribution system.

Transmission pipelines are establishments primarily engaged in the pipeline transportation of natural gas from gas fields or processing plants to local distribution systems.

E - Receipts from Storage Facilities

Report volumes of gas received from storage facilities (NAICS 493190) connected directly to your company's distribution system.

Storage facilities include natural gas storage caverns and liquefied natural gas storage, but exclude establishments primarily engaged liquefaction and regasification of natural gas for purposes of transport (NAICS 488990).

F - Receipts from Other Gas Distributors

Report volumes of gas received from other gas distributors (NAICS 221210) connected directly to your company's distribution system.

Gas distributors are establishments primarily engaged in the distribution of natural or synthetic gas to the ultimate consumers through a system of mains.

G - Total Supply of Natural Gas

Report total volumes of gas received.

H - Average Heating Value in Gigajoules/Thousand Cubic Meters

Report average heat content of your natural gas receipts for the reported reference month.

Disposition

I - Deliveries to System Gas Consumers

Report deliveries of utility-purchased natural gas to consumers. Report the quantity and value of the natural gas delivered and the number of customers.

J - Deliveries to Consumers Enrolled with a Third Party Marketer

Report deliveries to consumers who have purchased their natural gas through a gas marketer or broker. Report the quantity and value of natural gas delivered and the number of customers.

K - Deliveries to Consumers who have Purchased Directly from Suppliers

Report deliveries to consumers who have purchased their natural gas directly from suppliers. Report the quantity of natural gas delivered and the number of customers.

L - Deliveries to Power Generation Plants

Report gas delivered to electric power generation plants (NAICS 2211) connected directly to your company's distribution system (at metered interconnections).

This industry comprises establishments primarily engaged in the generation of bulk electric power by natural gas.

M - Deliveries to Other Industrial Consumers

Report gas delivered to industrial establishments other than power generation plants.

Inclusions:

• Agriculture and forestry
• Mining, quarrying, and oil and gas extraction
• Construction
• Manufacturing

Exclusions:

• Electric power generation
• Wholesale and retail trade
• Transportation and warehousing
• Other commercial buildings (e.g., public institutions)
• Natural gas transmission pipelines
• Natural gas storage facilities
• Natural gas distributors

N - Deliveries to Commercial and Institutional Consumers

Report gas delivered to commercial and institutional establishments.

Inclusions:

• Wholesale and retail trade
• Transportation and warehousing
• Other commercial buildings (e.g., public institutions)

O - Deliveries to Residential Consumers

Report gas delivered for domestic use (including multi-dwelling apartments).

P - Deliveries to Transmission Pipelines

Report volumes of gas delivered to transmission pipelines (NAICS 486210) connected directly to your company's distribution system.

Transmission pipelines are establishments primarily engaged in the pipeline transportation of natural gas from gas fields or processing plants to local distribution systems.

Q - Deliveries to Storage Facilities

Report volumes of gas delivered to storage facilities (NAICS 493190) connected directly to your company's distribution system.

Storage facilities include natural gas storage caverns and liquefied natural gas storage but exclude establishments primarily engaged liquefaction and regasification of natural gas for purposes of transport (NAICS 488990).

R - Deliveries to Other Gas Distributors

Report volumes of gas deliveries to other gas distributors (NAICS 221210) connected directly to your company's distribution system.

Gas distributors are establishments primarily engaged in the distribution of natural or synthetic gas to the ultimate consumers through a system of mains.

S - Own Use

Report volumes of gas consumed in operating your pipeline system.

T - Line Pack Fluctuation

Report differences in the pipeline system due to changes of temperature and/or pressure.

U - Metering Differences, Line Loss, Other Unaccounted for and Cyclical Billing Adjustments

Report the difference between the total supply and total disposition. This difference includes leakage or other losses, discrepancies due to meter inaccuracies and other variants, particularly billing lag.

V - Average Heating Value in Gigajoules/ Thousand Cubic Meters

Report the average heat content of your total natural gas disposition for the reference month.

W - Total Disposition

Report total volumes of gas disposition.

Thank you for your participation.

Integrated Business Statistics Program (IBSP)

This guide is designed to assist you as you complete the 2025 Biannual Potato Area and Yield Survey - June. If you need more information, please call the Statistics Canada Help Line at the number below.

Your answers are confidential.

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act.

Statistics Canada will use information from this survey for statistical purposes.

Help Line: 1-877-949-9492 or TTY 1-855-382-7745

Table of contents

• A - Reporting instructions
• B - Definitions
• C - Question 1 and 2

A - Reporting instructions

• When precise figures are not available, please provide your best estimates.

B - Definitions

Legal Name

The legal name is one recognized by law, thus it is the name liable for pursuit or for debts incurred by the business or organization. In the case of a corporation, it is the legal name as fixed by its charter or the statute by which the corporation was created.

Modifications to the legal name should only be done to correct a spelling error or typo.

To indicate a legal name of another legal entity you should instead indicate it in question 3 by selecting 'Not currently operational' and then choosing the applicable reason and providing the legal name of this other entity along with any other requested information.

Operating Name

The operating name is a name the business or organization is commonly known as if different from its legal name. The operating name is synonymous with trade name.

Current main activity of the business or organization

The North American Industry Classification System (NAICS) is an industry classification system developed by the statistical agencies of Canada, Mexico and the United States. Created against the background of the North American Free Trade Agreement, it is designed to provide common definitions of the industrial structure of the three countries and a common statistical framework to facilitate the analysis of the three economies. NAICS is based on supply-side or production-oriented principles, to ensure that industrial data, classified to NAICS, are suitable for the analysis of production-related issues such as industrial performance.

The target entity for which NAICS is designed are businesses and other organizations engaged in the production of goods and services. They include farms, incorporated and unincorporated businesses and government business enterprises. They also include government institutions and agencies engaged in the production of marketed and non-marketed services, as well as organizations such as professional associations and unions and charitable or non-profit organizations and the employees of households.

The associated NAICS should reflect those activities conducted by the business or organizational unit(s) targeted by this questionnaire only, and which can be identified by the specified legal and operating name. The main activity is the activity which most defines the targeted business or organization's main purpose or reason for existence. For a business or organization that is for-profit, it is normally the activity that generates the majority of the revenue for the entity.

The NAICS classification contains a limited number of activity classifications; the associated classification might be applicable for this business or organization even if it is not exactly how you would describe this business or organization's main activity.

Please note that any modifications to the main activity through your response to this question might not necessarily be reflected prior to the transmitting of subsequent questionnaires and as a result they may not contain this updated information.

C - Question 1 and 2

Are you growing any potatoes for sale this year? 

What is the total area of potatoes planted in this crop year?

Please report for the entire operation. Report the area of potatoes planted on land owned or rented by all partners in the operation.

Please report all planting intentions, if you have not completed your planting activities when completing this survey.

Crop Year

The period of time between one year's harvest to the next.

For most provinces, the crop year is from August to the following July.

However, in British Columbia, they could harvest potatoes as early as June so their crop year could run from June to the following May.

Thank you for your participation.

By Pierre Zwiller-Panicz, Margarita Novikova, Kirsten Gaudreau, Matthew Paslawski

Summary

This study develops and implements a machine learning model to forecast expenditures in HICC Grants and Contributions programs, focusing on reimbursement-based claims. A comparative analysis of algorithms identified Random Forest as the most effective, achieving an R-squared of 39%. Integrated into a Power BI dashboard, the model enables real-time expenditure analysis, trend visualization, and actual vs. forecast comparisons. Its implementation reduced forecasting time from three months to one, improving financial planning and stakeholder engagement.

The model has had a significant business impact, streamlining discussions between financial managers (FMAs) and program stakeholders while providing real-time insights for better decision-making. Although its applicability is limited to programs with established projects and it performs less accurately for allocation-based programs, it has proven highly effective for reimbursement-based claims with 30 or more active projects.

With its demonstrated success, the model represents a valuable step forward in financial forecasting. Its implementation has set the stage for further advancements, paving the way for broader adoption and continued improvements in predictive accuracy and program applicability.

1. Introduction

Housing, Infrastructure, and Communities Canada plays a crucial role in funding and supporting infrastructure projects that contribute to sustainable, inclusive, and climate-resilient communities. The department’s G&C programs require detailed, multi-year financial forecasts to ensure efficient allocation of government funding. However, the unpredictable nature of infrastructure projects often leads to overstated cashflow estimates, resulting in unspent funds and budget inefficiencies. As HICC's G&C programming continues to expand, the need for a scalable, data-driven forecasting solution has become increasingly clear.

To address these challenges, HICC implemented a ML forecasting model in May 2024. This innovative tool leverages advanced analytics to improve expenditure forecasting, enhance financial planning accuracy, and optimize budget allocation. By integrating this model into HICC’s existing suite of forecasting tools, the department aims to reduce inefficiencies, support data-driven decision-making, and strengthen its ability to fund critical infrastructure initiatives.

This article explores the development and implementation of the ML G&C forecasting model. It begins with an overview of the project’s background and objectives, followed by the model’s technical development and integration into HICC’s financial forecasting processes. The results and their impact on financial planning are then analyzed. The article concludes with recommendations for future enhancements and potential applications of the model.

2. ML Forecasting Model Background

2.1. Background and Evolution of Initiatives

In fiscal year 2016-2017 and 2017-2018, HICC lapsed approximately 64% of its anticipated Grants and Contributions authorities which led to a demand by central agencies for predictability in forecasting the fiscal profile of infrastructure programming. In response, HICC took several steps toward addressing these challenges:

• 2019–2020: HICC established a departmental Tiger Team to examine all aspects of the management of contribution funding to better align appropriations with actual expenditures.
• 2020–2022: HICC established a Grants and Contributions Centre of Expertise with a finance-focused skill set to address these issues.

2.2. Challenges in Current G&C Forecasting

Since its inception, the GCCOE has developed a suite of forecasting methods and processes that has helped reduce the departmental lapse in Grants Contribution funding. However, these methods and processes have created extensive workload on Financial Management Advisors (FMAs) to generate accurate forecasts which lacked standardization between different programs.

2.3. Purpose and Objective of the Model

To address these gaps, the GCCOE partners with the OCDO to explore a data-driven approach which could complement and build upon HICC’s existing suite of forecasting methods by providing a more accurate basis for FMA forecasts while simultaneously reducing workload.

The ML model's key objective is to improve forecasting of G&Cs at HICC by developing an automated forecasting tool informed by historical G&C data that is adaptable to current and new programming, thereby improving efficiency of HICC’s G&C multi-year forecasting process.

3. ML Forecasting Model Development & Implementation

This section will explore the development of the ML forecasting model, from its data sources to the final interactive tool. It will detail how the model was designed, integrated, and deployed to provide FMAs with real-time insights.

3.1. Data Sources

The first step in developing the ML Forecasting Model involved sourcing data from HICC’s Program Information Management System (PIMS), as illustrated in Figure 1 – Data Pipeline. PIMS provided detailed information on program funding and expenditures across three layers: Programs, Contribution Agreements, and Projects. Key variables included:

Variable	Definition	Sample Data
Project ID	A unique identifier for each project	13176
Contribution Agreement ID	A unique identifier linking the project to a specific funding agreement.	2
Fiscal Year	The fiscal year associated with the project’s expenditures and cashflows.	2007-2008
Project Cashflow	The projected or actual cash inflow and outflow for the project.	500 000
Project Expenditure	The amount spent on the project within a given period.	500 000
Total Amount per Contribution Agreement	The full allocated budget under a specific agreement.	2 000 000
Total Amount per Program Contribution	The overall funding contribution assigned to the program, encompassing multiple agreements.	2 000 000
Project Status	Indicates the current state of the project (e.g., active, completed, pending).	Closed

3.2. Data Preprocessing

3.2.1. Cleansing and Transformation

The data cleaning process began by identifying and removing blank entries that were irrelevant to the financial forecasting model. The final dataset included only projects with statuses—"closed," "completed," and "in implementation"—ensuring a comprehensive evaluation across all stages of the project lifecycle, which enhanced the model's robustness and adaptability.

Next, data manipulation was performed to generate key variables such as average expenditure, prior expenditures, remaining amounts, and contribution agreement values. Finally, a disaggregation process standardized the data to a consistent level of granularity. Initially structured at multiple levels—project, contribution agreement, and program—the dataset was ultimately consolidated at the project level to align with the forecasting model’s analytical framework.

To enhance the model's ability to capture financial constraints and monitor spending limits, several engineered features were derived from existing variables. These features include the total project amount, cumulative project amount, cumulative previous expenditures, recent expenditures, project lifecycle, average preceding expenditures, remaining funds, and remaining funds at the start of each fiscal year. By integrating these variables, the dataset was enriched with additional financial insights, ensuring a more accurate representation of project spending dynamics.

New Variables	Definition
Total Project Amount	The total budget allocated for a project over its entire duration. This is the sum of all planned expenditures for the project.
Cumulative Project Amount	The total amount spent on the project from its start up to the current fiscal year. This helps track how much of the budget has been utilized.
Cumulative Previous Expenditures	The sum of expenditures from all previous years before the current fiscal year. This excludes spending in the current year but provides historical financial context.
Recent Expenditures	The expenditures from the most recent fiscal year, reflecting the latest spending trends.
Project Lifecycle	The total number of years the project is expected to run, from its start year to its planned completion.
Average Preceding Expenditures	The average amount spent per year in past years. This is calculated as Cumulative Previous Expenditures / (Current Year - Start Year).
Remaining Funds	The total project budget minus cumulative expenditures. This represents the funds still available for future spending.
Remaining Funds at the Start of Each Fiscal Year	The amount of money left unspent at the beginning of a new fiscal year, before any new expenditures are incurred.
Amount	A derived variable used to improve forecast accuracy. Since future expenditures are initially 0, the model tends to predict unrealistically low values. Instead, the Amount variable replaces missing future expenditures with Cashflow (expected future disbursements), while keeping past expenditures unchanged.

3.2.2. Segmentation

With the cleaned and transformed dataset in place, the next step involved analyzing the distribution of project expenditure to better inform the modeling approach. As illustrated in Figure 2 – Project Amount Distribution, the dataset exhibited significant heterogeneity, with 95% of projects accounting for only 5% of the department’s total dollar contribution, while the remaining 5% represented 95% of the expenditures.

Given the significant disparities in project expenditures our finance colleagues initially recommended a segmentation approach. Their manual classification was based on project amounts to account for this imbalance. To refine this approach, we explored an advanced segmentation methodology. Instead of segmenting by project amount, our analysis suggested that project duration provided a better differentiation because, when comparing results, project duration offered better homogeneity within clusters, leading to more consistent groupings and improved predictive performance. During the development phase, shifting the segmentation criterion from total project cost to project duration reduced the MAE for the Random Forest model by at least 300,000. This initial categorization of projects into high materiality (>5 years) and low materiality (<5 years) led to a measurable improvement in model performance.

3.2.3. Principal Component Analysis (PCA)

To address multicollinearity among key variables, we incorporated Principal Component Analysis (PCA). Variance Inflation Factor analysis revealed severe collinearity, as illustrated in Figure 3, particularly among financial variables such as Project Total Amount, Cumulative Project Amount, and Remaining. This redundancy posed a risk of distorting predictions, especially for large-scale programs like Investing in Canada Infrastructure Program (ICIP)^{Footnote 1}.

By applying PCA, we transformed the original features into orthogonal components, capturing the maximum variance in a reduced-dimensional space. The explained variance analysis showed that five components retained about 90% of the total variance, preserving most of the dataset's information while reducing dimensionality. This trade-off mitigates multicollinearity while maintaining the predictive power of key features.

The PCA was used to reduce multicollinearity among the financial variables while preserving the most informative features. Figure 5: PCA results illustrates this method loadings, which represent how strongly each original variable contributes to a given principal component.

To improve interpretability, the principal components were renamed based on their dominant loadings:

1. Principal Component 1: Project Funding & Scale – This component is influenced by TOTAL_PROGRAM_CONTRIBUTION_AMT (0.41), project_total_amount (0.41), and cumulative_project_amount (0.37). It represents the overall financial scale of a project, emphasizing the total funding available.
2. Principal Component 2: Remaining Resources – This component captures the availability of unspent funds, primarily driven by Remaining (0.46) and Remaining_start_year (0.41). It indicates the funding is still accessible for ongoing projects.
3. Principal Component 3: Initial Resources – This component is moderately influenced by TOTAL_PROGRAM_CONTRIBUTION_AMT (0.21), project_total_amount (0.10), and cumulative_project_amount (0.04), suggesting it relates to the initial allocation of financial resources at the start of a project.
4. Principal Component 4: Project Duration – This component strongly correlates with project_lifetime (0.70), indicating that it captures project longevity and its relationship to past spending trends.
5. Principal Component 5: Lifecycle Contributions – This component captures the financial balance across a project’s timeline, with strong contributions from project_lifetime (0.61) and previous_spend (0.26).

By integrating PCA into our modeling pipeline, we effectively addressed the collinearity issues present in the original dataset and improved the stability and interpretability of the model.

The review also highlighted an important consideration: if most of the variance is not captured within a few components, it may indicate a complex data structure or non-linear relationships. In such cases, techniques like Kernel PCA, t-SNE, or UMAP might be more suitable. However, since PCA with five components retains 90% of the variance, it remains a valid choice for dimensionality reduction in this context. Future work could further explore non-linear embeddings to determine if an alternative approach provides additional performance gains.

4. G&Cs Machine Learning Forecasting Model Development

With the pre-processing complete, the next phase focused on building a robust forecasting model. This involved selecting an appropriate algorithm, fine-tuning hyperparameters, and evaluating performance to ensure accuracy across diverse project scales. Given the complexity of financial data, our approach prioritized interpretability, stability, and alignment with business needs.

4.1. Final Dataset

Variable	Definition	Sample Data
Project ID	A unique identifier for each project	13176
Contribution Agreement ID	A unique identifier linking the project to a specific funding agreement.	2
Fiscal Year	The fiscal year associated with the project’s expenditures and cashflows.	2007-2008
Project Cashflow	The projected or actual cash inflow and outflow for the project.	500 000
Project Expenditure	The amount spent on the project within a given period.	500 000
Total Amount per Contribution Agreement	The full allocated budget under a specific agreement.	2 000 000
Project Status	Indicates the current state of the project (e.g., active, completed, pending).	Closed
Amount	A derived variable used to improve forecast accuracy. Target variable.	500 000
PC1	Project funding & scale: It represents the overall financial scale of a project, emphasizing the total funding available.	-0.68
PC2	Remaining resources: It indicates the funding is still accessible for ongoing projects.	0.97
PC3	Initial resources: it relates to the initial allocation of financial resources at the start of a project.	-1.34
PC4	Project Duration: it captures project longevity and its relationship to past spending trends.	-0.19
PC5	Lifetime Contribution: captures the financial balance across a project’s timeline, with strong contributions from project_lifetime and previous_spend	-0.08

Note: As segmentation was part of the test scenarios, we initially maintained two separate datasets—df_high and df_low—grouping projects based on their materiality level.

4.2. Model Training

The dataset was structured as a time series, covering fiscal years from 2003-2004 up to 2023-2024. It was split into a training set (75%) and a test set (25%), ensuring that past data was used to forecast future expenditures. Once trained, the model was applied to predict expenditures for the current fiscal year (2024-2025) and projected three years into the future. The training process was iterative, refining the models to optimize performance while maintaining stability.

4.3. Model Comparison

Several models were evaluated, including Random Forest, Gradient Boosting, and XGBoost, based on their predictive accuracy and ability to capture patterns in financial data. Since expenditures follow a sequential pattern over time, the models needed to account for temporal dependencies and underlying trends.
Each model had distinct characteristics:

• Random Forest, an ensemble method, effectively captured complex interactions, making it a strong candidate for financial forecasting.
• Gradient Boosting refined predictions through iterative learning, improving accuracy.
• XGBoost, an optimized gradient boosting algorithm, enhanced accuracy further and addressed overfitting concerns.

Model performance was evaluated using two key metrics:

• R² (Coefficient of Determination): Measures how well the model explains variance in expenditures.
• MAE (Mean Absolute Error): Quantifies the average prediction error, providing a clear financial accuracy measure.

4.4. Model Performance Assessment

This section presents the assessment metrics used to compare different models. The objective was to balance predictive accuracy, stability, and interpretability while capturing the complexities of financial data.

Scenarios	Characteristics	Best Model Performance (metrics)
Scenario 1	No segmentatio PCA	Random Forest (MAE: 137,570; R²: 93%) Overfit
Scenario 2	No segmentation PCA cap outliers	Random Forest (MAE: 852,243; R²: 36%)
Scenario 3	No segmentatio PCA cap outliers n_estimators=100 max_depth=10 min_samples_leaf=5 random_state=4	Random Forest (MAE: 888,558; R²: 37%)
Scenario 4	No segmentation PCA cap outliers weights for closer years n_estimators=100 max_depth=10 min_samples_leaf=5 random_state=42	Random Forest (MAE: 888,526; R²: 81%)
Scenario 5	No segmentation PCA cap outliers automated hyperparameter tuning	Random Forest (MAE: 758,012; R²: 40%)

4.5. Model Performance Assessment Consideration

To optimize model performance, various test splits were evaluated, including both 25%-30% and automated splits. Each scenario was tested to assess how different training and testing data partitions impacted the model's accuracy and generalizability. The automated split approach was also considered, ensuring the model’s performance remained robust across different data partitioning strategies. This comprehensive testing allowed for the identification of the most effective split configuration to enhance the model's predictive capabilities.

After optimizing the test split, we evaluated the impact of segmentation and PCA on model performance. While segmentation was initially expected to improve accuracy, testing showed that removing it while retaining PCA resulted in more stable and precise forecasts. Initially, segmentation led to the creation of two separate datasets, grouping projects based on predefined criteria. However, further analysis revealed that a unified dataset provided greater consistency, simplified model implementation, and improved scalability. This finding underscored the effectiveness of dimensionality reduction in capturing key patterns while minimizing noise. Consequently, PCA was prioritized as the primary technique for managing data complexity, ensuring a more robust and generalizable forecasting model.

Finally, we conducted hyperparameter tuning using both manual adjustments and automated optimization. However, tuning did not always yield significant improvements. For instance, a tuned Random Forest model achieved a slightly lower MAE (811,547) and a higher R² (39%), but this marginal gain did not justify the increased complexity. In some cases, hyperparameter tuning introduced overfitting, reinforcing the reliability of the simpler, default configuration.

4.6. Best Model Results

Our analysis revealed that Random Forest consistently outperformed other models in balancing predictive accuracy and alignment with financial forecasts. Among the tested configurations, the best-performing model was Random Forest (MAE: 852,243, R²: 36%), which provided expenditure projections of 4.8B for 2024-2025 and 5.2B for 2025-2026, making it the closest to GCCOE predictions. The formula for accuracy is:

4.7. ML Forecast Model Limitations

Despite the promising results, there are several limitations to the ML forecasting model that need to be addressed. Programs without project records in the system cannot be modeled because the model needs to know a project exists to generate a forecast for it. In addition, the model was created to forecast reimbursement-based claims for direct delivery programs. As such, it is less accurate when it comes to forecasting for other types of payments such as grants, advances, and milestone-based claims. The model is also less accurate for allocation-based or flow through G&C programs. Lastly, the model’s accuracy diminishes at the individual project level, which can exhibit atypical spending patterns.

These limitations mean the model currently performs best for direct delivery programs with 30 or more active projects in the system and for which the majority of claims are reimbursement based.

5. Business Results

5.1. How to get ML in the hands of the business:

To enhance the interpretation of the model's results, forecasted expenditures were integrated into an existing dashboard used by FMAs (Figure 7). This dashboard visually represents expenditure trends, allowing comparisons between actual expenditures, the FMAs' current manual method, and the ML model's forecasts.

The interactive dashboard also features personalized report capabilities, allowing users to customize their views by selecting specific criteria, such as province, program, and fiscal year. This flexibility ensures that users can tailor the data exploration to meet their specific analytical needs. Furthermore, the interactive nature of the visualization allows users to hover over any point on the curves to view exact values for each fiscal year, offering a more intuitive and granular exploration of the data. This combination of dynamic reporting and interactive visualizations supports in-depth analysis and facilitates decision-making based on the ML model's output.

5.2. Business Impact

The ML model was implemented in May 2024 to forecast G&C spending for fiscal years 2024-25 through 2027-28. It generated multi-year forecasts for nine of the department’s G&C programs, which account for approximately 80% of the department’s Grants and Contributions funding. The model's accuracy will be evaluated in April 2025 and April 2026, at the close of fiscal years 2024-25 and 2025-26, respectively.

The model’s implementation streamlined the forecasting process, reducing the time required from three months to one month. This was achieved by providing FMAs with a baseline forecast – generated by the ML model – which facilitated discussions with their respective programs and aligned expectations ahead of the recipient cash flow collection process.

Additionally, the integrated dashboard supports ongoing discussions with stakeholders, using up-to-date data in preparation for regular departmental reporting.

6. Conclusion and Next Steps

In conclusion, this study highlights the significant potential of the implementation of a ML forecasting model in the context of expenditure prediction for HICC G&C programs. The model demonstrates a high level of accuracy when compared against historic expenditures and is being tested against actual expenditures over the next two years in hopes of optimizing Grants and Contributions funding, reducing public account lapses, and streamlining financial processes. Despite the challenges and limitations mentioned, the overall results have shown promise in enhancing financial decision-making and operational efficiency.

The success of this initiative was formally recognized in December 2024 when the project received the 2024 Controller General Innovation Award, recognizing its significant impact in financial management. Since then, the model has garnered attention across various government departments, prompting consultations to explore its broader application. The ongoing efforts to promote its adoption reflect a growing recognition of the potential for ML-driven solutions to improve financial forecasting and resource allocation across the public sector.

Additionally, the project was longlisted for the 2025-2026 edition of the Public Service Data Challenge. This recognition highlights the increasing interest and enthusiasm from multiple departments to adopt the ML-driven forecasting tool. Ongoing efforts to promote its adoption further underscore the growing recognition of the potential for machine learning solutions to enhance financial forecasting and optimize resource allocation across the public sector.

If you have any questions about my article or would like to discuss this further, I invite you to Meet the Data Scientist, an event where authors meet the readers, present their topic, and discuss their findings.

Register for the Meet the Data Scientist event. We hope to see you there!

Subscribe to the Data Science Network for the Federal Public Service newsletter to keep up with the latest data science news.