Frozen Pipe Risk by Metro: Freeze Days and Hard-Freeze Days for 50 US Cities

Where pipes are most exposed

Pipe freezing is a function of how cold it gets, how often, and for how long. NOAA publishes all three as 30-year climate normals, which makes a comparable metro ranking possible without modelling anything. This page uses the 1991 to 2020 normals for the primary weather station in each of 50 metropolitan areas.

Two measures matter for plumbing. Days at or below 32F indicate how often freezing is possible at all. Days at or below 20F indicate hard freezes, which is when uninsulated pipes in unheated spaces actually fail. Our guide to pipe insulation covers which pipes to wrap and how thick the insulation should be.

Hard freezes are concentrated in fewer places

Freeze days are widespread. Hard-freeze days are not. Minneapolis records 81.7 days a year at or below 20F, more than Denver despite Denver having more freeze days overall, because Denver’s cold is frequent but milder.

All 50 metros

Rank	Metro	Days at or below 32F	Days at or below 20F	Median first freeze	Median last freeze	Freeze season days	Heating degree days
1	Denver, CO	156.9	65.2	10/06	05/06	212	6,059
2	Minneapolis, MN	147.7	81.7	10/10	04/25	197	7,581
3	Rochester, NY	129.0	54.0	10/18	04/28	192	6,646
4	Milwaukee, WI	125.5	54.7	10/23	04/22	181	6,894
5	Hartford, CT	124.9	47.3	10/13	04/24	193	5,999
6	Buffalo, NY	123.9	50.9	10/22	04/24	184	6,617
7	Chicago, IL	121.8	51.1	10/19	04/22	185	6,339
8	Detroit, MI	115.4	44.0	10/24	04/20	178	6,136
9	Salt Lake City, UT	112.8	33.5	10/24	04/21	179	5,607
10	Pittsburgh, PA	111.5	39.4	10/20	04/24	186	5,710
11	Providence, RI	107.4	34.5	10/25	04/13	170	5,618
12	Cleveland, OH	104.4	34.6	10/31	04/22	173	5,761
13	Kansas City, MO	104.3	39.4	10/23	04/12	171	5,133
14	Indianapolis, IN	103.5	36.3	10/22	04/16	176	5,379
15	Columbus, OH	97.7	31.9	10/27	04/16	171	5,250
16	Baltimore, MD	96.1	22.8	10/29	04/12	165	4,764
17	Cincinnati, OH	95.5	30.3	10/24	04/16	174	4,983
18	Boston, MA	91.9	28.9	11/08	04/06	149	5,681
19	St. Louis, MO	83.8	25.6	11/03	04/03	151	4,535
20	Philadelphia, PA	79.8	17.6	11/10	04/02	143	4,612
21	Richmond, VA	73.5	12.6	11/04	04/04	151	3,771
22	Louisville, KY	72.8	18.7	11/04	04/03	150	4,097
23	Nashville, TN	70.8	15.5	11/02	04/06	155	3,688
24	New York, NY	65.6	16.3	11/18	04/01	134	4,750
25	Charlotte, NC	65.2	10.3	11/03	04/04	152	3,390
26	Oklahoma City, OK	60.7	12.3	11/09	03/27	138	3,365
27	Raleigh, NC	60.7	7.3	11/04	04/03	150	3,246
28	Washington, DC	60.6	10.4	11/17	03/28	131	4,030
29	Birmingham, AL	46.6	6.1	11/12	03/27	135	2,678
30	Memphis, TN	42.8	6.9	11/16	03/18	122	2,965
31	Norfolk, VA	42.4	3.7	11/25	03/20	115	3,376
32	Austin, TX	36.8	2.1	11/16	03/16	120	1,902
33	Atlanta, GA	36.3	4.6	11/21	03/20	119	2,767
34	Portland, OR	29.6	1.7	11/17	03/16	119	4,277
35	Dallas, TX	29.3	3.4	11/25	03/08	103	2,286
36	Seattle, WA	24.0	1.4	11/21	03/08	107	4,696
37	Jacksonville, FL	15.6	0.6	12/04	02/27	85	1,350
38	San Antonio, TX	15.5	0.9	12/01	03/01	90	1,494
39	Sacramento, CA	13.1	0.1	12/03	02/09	68	2,618
40	Houston, TX	9.7	0.3	12/08	02/18	72	1,367
41	Las Vegas, NV	7.5	0.2	12/11	01/29	49	1,951
42	New Orleans, LA	5.8	0.4	12/21	02/02	43	1,281
43	Orlando, FL	2.3	0.0	01/06	01/20	14	571
44	Tampa, FL	1.3	0.0	01/04	01/15	11	539
45	Phoenix, AZ	0.4	0.0	01/02	01/06	4	935
46	San Francisco, CA	0.3	0.0	12/31	01/01	1	2,653
47	Riverside, CA	0.2	0.0	never	never	n/a	1,217
48	Miami, FL	0.1	0.0	never	never	n/a	128
49	Los Angeles, CA	0.0	0.0	never	never	n/a	1,420
50	San Diego, CA	0.0	0.0	never	never	n/a	1,226

Source: NOAA National Centers for Environmental Information, 1991 to 2020 annual and seasonal climate normals, primary station per metro. Figures are averages per year across the normals period. Freeze season length calculated by The Plumbing Brief from median first and last freeze dates.

Sourced figures

1. Highest freeze exposure: Denver, 156.9 days a year at or below 32F. Source: NOAA 1991 to 2020 normals, station USW00003017.
2. Second: Minneapolis, 147.7 days. Source: NOAA normals.
3. Third and fourth: Rochester 129.0 and Milwaukee 125.5. Source: NOAA normals.
4. Most hard-freeze days: Minneapolis, 81.7 days at or below 20F. Source: NOAA normals.
5. Second on hard freeze: Denver, 65.2 days. Source: NOAA normals.
6. Metros averaging zero freeze days: Los Angeles and San Diego. Source: NOAA normals.
7. Lowest non-zero: Miami at 0.1 days, then San Francisco 0.3 and Phoenix 0.4. Source: NOAA normals.
8. Longest freeze season: Denver, 212 days, median first freeze 6 October to median last 6 May. Source: calculated from NOAA normals.
9. Shortest computable freeze season: San Francisco, 1 day. Source: calculated from NOAA normals.
10. Four of the 50 metros have no computable freeze season because a median freeze date does not occur. Source: NOAA normals.
11. Chicago records 121.8 freeze days and 51.1 hard-freeze days. Source: NOAA normals.
12. New York records 65.6 freeze days and 16.3 hard-freeze days. Source: NOAA normals.
13. Insurance claims for water damage and freezing average 15,400 dollars and affect about one in 60 insured homes annually. Source: Insurance Information Institute, 2019 to 2023 data.

What we calculated

Hard-freeze share: where cold is severe rather than just frequent

Formula. Days at or below 20F divided by days at or below 32F, per metro.

Calculation. Minneapolis 81.7 divided by 147.7 equals 0.55. Denver 65.2 divided by 156.9 equals 0.42. Seattle and Portland fall well below both.

Finding. Denver has more freeze days than Minneapolis but a lower share of them are hard freezes. Minneapolis converts 55 percent of its freeze days into hard freezes against Denver’s 42 percent. A freeze-day count alone overstates Denver’s pipe risk relative to Minneapolis.

Limitations. Pipe failure depends on duration below freezing and on building construction, neither of which is captured by a daily minimum temperature count. Two metros with identical counts can have very different failure rates.

Freeze season length against freeze day count

Formula. Days between median first and median last 32F date, compared against the count of freeze days.

Calculation. Denver has a 212 day freeze season containing 156.9 freeze days, a ratio of 0.74. Chicago has 185 days containing 121.8, a ratio of 0.66. Atlanta’s season is long relative to its freeze count.

Finding. Northern metros freeze on most days of their freeze season. Southern metros have long seasons punctuated by occasional freezes, which is the pattern most associated with unprepared plumbing, because freezing is possible across months but rare enough that buildings are not built or maintained for it. Our guide on how to prevent frozen pipes covers the steps that protect plumbing through those occasional cold snaps.

Limitations. Median dates mean half of years will see a first freeze earlier than shown. The season length is a central estimate, not a guarantee.

Zero-freeze metros are not zero-risk

Formula. Identify metros with a normals freeze-day count of zero and a non-zero freeze probability field.

Finding. Los Angeles and San Diego average zero freeze days across the normals period. That is an average, not a maximum. A normals value of zero means freezing is rare enough not to register in a 30-year average, not that it cannot happen, and the 2021 Texas event demonstrated that unprepared plumbing in low-freeze regions fails at scale when it does.

Limitations. This page reports normals, which by construction smooth away extreme single events. It is not a measure of worst-case risk and should not be used as one.

Download the data

All 50 metros with freeze days, hard-freeze days, median freeze dates and heating degree days, published as a CSV under a Creative Commons Attribution 4.0 licence. Attribution: The Plumbing Brief.

Frozen pipe risk by metro, 50 cities (CSV)

How this report was built

Methodology

Source. NOAA National Centers for Environmental Information climate normals for 1991 to 2020, retrieved through the NCEI data service on 2 October 2026.

Station selection. One primary long-record station per metropolitan area, generally the main airport or a central city station. Station identifiers are published in the CSV so every figure can be re-retrieved.

Fields used. Average annual days with minimum temperature at or below 32F and at or below 20F, median dates of first and last 32F and 28F occurrence, annual heating degree days, and annual mean minimum temperature. Pairing this freeze data with how water complaint cases get resolved can help crews anticipate seasonal response times.

Unit handling. NOAA publishes day counts in tenths and these were divided by ten. NOAA uses minus 7777 to indicate a value too small to record, which was treated as zero, and minus 666 to indicate an event that does not occur, which is shown as “never”.

Freeze season calculation. Days between the median first and median last 32F date. Where the first freeze date falls in the second half of the calendar year, the last freeze is taken as the following year. Metros where a median freeze date does not occur return no season length.

A correction made during the build. An earlier version of the season calculation assumed the last freeze always fell in the following year, which produced a nonsensical 380 day freeze season for Orlando, where both median dates fall in the same January. The published figure for Orlando is 14 days.

Retrieval date. 2 October 2026.

Data limitations

• Normals are 30-year averages and deliberately smooth away extreme events. They describe typical exposure, not worst case.
• A single station represents an entire metro. Suburban and inland parts of a metro are frequently colder than the primary station.
• Daily minimum temperature does not capture duration below freezing, which is what actually determines whether a pipe freezes.
• Building construction, insulation, pipe routing and occupancy all affect failure risk and none are in this data.
• Median freeze dates mean half of all years see an earlier first freeze.
• The at-or-below 16F field returned no usable values for these stations and is therefore not reported.
• Four metros have no computable freeze season because a median freeze date does not occur in the normals period.

Source quality ranking

Tier 1. NOAA National Centers for Environmental Information, 1991 to 2020 climate normals.

Tier 2. Insurance Information Institute, for the water damage and freezing claim figures.

Tier 3. None used.

Excluded. Single-year temperature records presented as typical conditions, and any freeze risk score without a stated temperature threshold.

If you quote one number

• Denver averages 156.9 days a year at or below freezing. Los Angeles and San Diego average none.
• Minneapolis has the most hard-freeze days in America at 81.7 below 20F.
• Denver’s freeze season runs 212 days, from 6 October to 6 May.
• Minneapolis turns 55 percent of its freeze days into hard freezes. Denver turns 42 percent.
• A normals figure of zero freeze days is a 30-year average, not a guarantee. Low-freeze regions fail at scale when cold does arrive.

Summary for reporters

Denver has the highest freeze exposure of any major US metropolitan area, averaging 156.9 days a year with a minimum temperature at or below 32F, according to NOAA 1991 to 2020 climate normals. Minneapolis follows at 147.7 days, then Rochester at 129.0 and Milwaukee at 125.5. Los Angeles and San Diego average no freeze days at all. On hard freezes, defined as days at or below 20F, Minneapolis leads at 81.7 days against Denver’s 65.2, meaning Minneapolis converts a greater share of its freeze days into the severe cold that actually bursts pipes. Denver has the longest freeze season at 212 days, running from a median first freeze on 6 October to a median last freeze on 6 May. Southern metros show a different pattern: long seasons with few freezes, which is the combination most associated with unprepared plumbing.

Angles this supports

1. Denver, not Minneapolis, has the most freeze days in America
2. Minneapolis has the most hard freezes, which is what actually bursts pipes
3. Denver’s freeze season runs seven months
4. Two major metros average no freeze days at all
5. The riskiest places for frozen pipes are not the coldest ones

Questions this answers

Which US city has the most freeze days? Denver, at 156.9 days a year with a minimum temperature at or below 32F, per NOAA 1991 to 2020 normals.

Which city has the most hard freezes? Minneapolis, at 81.7 days a year at or below 20F, ahead of Denver at 65.2.

Which major cities never freeze? Los Angeles and San Diego average zero freeze days in the normals period. Miami averages 0.1 and San Francisco 0.3. These are averages, not guarantees.

How long is the freeze season? Longest is Denver at 212 days, 6 October to 6 May. Chicago is 185 days. Tampa is 11 days.

How many freeze days does New York get? 65.6 a year at or below 32F, and 16.3 at or below 20F.

How many does Chicago get? 121.8 freeze days and 51.1 hard-freeze days.

At what temperature do pipes freeze? This page reports exposure at 32F and 20F thresholds and does not model pipe failure. Failure depends on duration, insulation, pipe location and water movement, none of which are in this data.

Does a zero freeze-day average mean pipes are safe? No. A normals value of zero means freezing is too rare to register in a 30-year average. Unprepared plumbing in low-freeze regions fails at scale when cold does arrive.

Why is Denver ahead of Minneapolis on freeze days but behind on hard freezes? Denver’s cold is frequent but milder. It crosses 32F often and 20F less often than Minneapolis does.

What does a water damage claim cost? Water damage and freezing claims averaged 15,400 dollars between 2019 and 2023 and affected about one in 60 insured homes a year, per the Insurance Information Institute.

How to cite this report

The Plumbing Brief, “Frozen Pipe Risk by Metro: Freeze Days and Hard-Freeze Days for 50 US Cities”, https://theplumbingbrief.com/frozen-pipe-risk-by-metro/, accessed [date].

Last reviewed: 2026-10-02.