Fire Alarm Design and AutoCAD Services

A few of the other services I provide are Printing, Web Design, Graphic Design and Business Implementation.

What I do.

I provide all of the necessary documents required for a submittal to the authority having jurisdiction for a new fire alarm system. I start with the design of the system using AutoCad 2011, and provide 3 ea. color copies of the new design using a HP designjet. I then create the voltage drop calculations, the standby battery calculations and any other calculations needed, such as speaker draw, using excel spreadsheets. All of the specification sheets are included for all of the devices used in the installation. A cover sheet is also provided showing a material list and quantities. I can be listed under your company so that I may sign the drawings with my APS stamp. More information can be found on my website at FireAlarmDesigns.com

Wednesday, February 9, 2011

AutoCAD Tutorials and Resource Library

The purpose of this blog is to provide valuable information and resources for commercial Fire Alarm design in the Fire Alarm industry. There will be posts on AutoCAD , .dwg files for download, calculators, voltage drop information, .pdf documents,etc...

A Letter from a local Fire Alarm authority concerning Voltage Drop

Lassen Wir Das Thema Voltage Fallen (let's voltage drop the subject).


As an apprentice, I read the National Fire Alarm Code only when I needed to look up bits and parts, a procedure here, or find a prescription there, like a cook book. I didn’t, and I suspect most beginners don’t browse the National Fire Alarm Code like a fishing catalog. I call this the JAW (Just Add Water) method. Such information is initially incomplete, and assumptions, inferences, and deductions thus drawn are imprecise and usually wrong.

The most common use of this document is to find expert to evidence to bolster a belief if not our ego. When that evidence is not found; it is the deliberate twisting of the intentions by manipulation of the language that causes confusion, not the document itself. On it’s face the document language is deliberately precise, relatively complete, and readily understood. That is to say, when we presume this thing or that as fact, and only then look to the document, we are using the standard the way a drunkard uses a lamppost, not for enlightenment, but for support.

Consider if you will a rejection notice from a city employee (note all references are from the 1996 edition NFC and remain unchanged in later editions)

‘Please revise and resubmit your voltage drop calculations. “Voltage drop calculations exceed 15% allowable line loss in accordance with NFPA 72, (National Fire Alarm Code) 1-5.5.1”.

The following 6 points leap to mind upon reading this statement;

1. Article 1-5.5.1 establishes a upper limit to a minimum operating parameter (voltage), it does not establish a maximum (voltage) loss parameter, and these two parameters are not reciprocal

2. The article 1-5.5.1 DOESN’T EVEN ALLUDE to voltage drop.

3. The number 15% is not enumerated in the article. In fact it article 1-5.5.1 has nothing to do with establishing a fixed value for voltage drop.

4. The NFC doesn’t have a need to (and thus would not) set a fixed value for voltage drop, for any device. That is done by the manufacturer. The NFC is intended to establish the required (minimum) levels of performance, not to establish the methods by which these requirements are to be achieved.

5. The NFC relies on the operating parameters set by the manufacturer and approved by the listing authority; and requires these are to be marked on the device itself..see 6-1.2)

6. Voltage drop specifications for circuits (if any) would be found in the National Electrical Code (NEC), as with all conductors. Not the NFC

So what made this person think that 1-5.5.1 specifies only a maximum 15% “Line Loss”? Well the NEC gives us a clue; it is a classic example of the third stage JAW method.

Voltage drop is quantified in three NFPA standards including the NEC (National Electrical Code), and where found; it is always expressed as a percentage.

None of those standards regulate or reference building fire alarm circuits.

Now here is what I think happens, follow my logic;

Being the AHJ and Upon reading section 907.1.1 (8) in the IBC 2000 edition I am in need of guidance (see below)

[F] SECTION 907

FIRE ALARM AND DETECTION SYSTEMS

907.1 General. This section covers the application, installation, performance and maintenance of fire alarm systems and their components.

907.1.1 Construction documents. Construction documents for fire alarm systems shall be submitted for review and approval prior to system installation. Construction documents shall include, but not be limited to, all of the following:

1. A floor plan.

2. Locations of alarm-initiating and notification

appliances.

3. Alarm control and trouble signaling equipment.

4. Annunciation.

5. Power connection.

6. Battery calculations.

7. Conductor type and sizes.

8. Voltage drop calculations.

Having heard the NEC limits the amount of voltage lost on branch circuits and feeders to the farthest outlet to 5% so I assume that Voltage drop is always express as a percentage (which is not true) and I look for the allowable percentage voltage drop for fire alarm circuits in the National Electrical Code. And I find nothing there not even in article 760 for fire alarm circuits. In the certain belief that if the building code requires voltage drop calculations there must be a standard that establishes their value I look to the National Fire Code in desperation and..“do de do dum de do… Ah Ha! “

“There it is!

the word voltage! right there in the title..

now what does that say .. yep! There’s the other word percentage that must be it, ..

but wait, ..oh this is tricky, it doesn’t actually say what the percentage of drop is.. it makes me cipher it. I’ll need my calculator, ..lets see 100% minus 85% is 15%, there! That must be it, the maximum allowed voltage drop is 15% of the name plate voltage.”

This is of course wrong

NFC 1-5.5.1 states..

“equipment shall be designed so that it is capable of performing its intended functions.. At 85 percent and at 110 percent of the nameplate primary (main) and secondary (standby) input voltage(s)”

That statement above, requires that any device connected to that circuit reliably operate even when the voltage supplied is a little higher or lower than the manufacturer expected, either on primary power or on (secondary) battery power; it does not say how much lower or higher. What it does say is; if you use a relay rated at 10 volts, it should be able to work even if the applied voltage is as little as 8.5 volts or as high as 11 volts. Now if your particular relay can operate on as little as 5 volts or as much as 20 volts, that’s good too, better even. Any way that’s still not voltage drop.

To understand we need to recognize what voltage drop is and then, why this is important;

Voltage drop is the difference in the voltage applied at one end and the voltage available on the other end. We need to know these values to be sure enough voltage remains on other end so that thing, whatever it is, will work the way it was designed even when operating on backup power. It is also important to note that 24 volt batteries often used as a secondary power source, are rated by their ability (Ah capacity) to support a given amount of current at a voltage of at least 20 volts for 20 hours. This means a tightly estimated system at the end of its designed standby time, could be operating at voltages as low as 20 volts when called upon to operate the appliance circuit. It is also important to note that when a manufacturer claims a minimum operating voltage of 20 volts the UL listing authority will require it to operate properly at voltages as low as 16 volts in order to be labeled with a rating of 20 volts.

The question then becomes a common sense one, and the NFC doesn’t attempt to regulate or standardize common sense (because it is so un-common).

The question should properly seeks to define the lowest voltage that must remain after voltage losses and that is of course simply the lowest (minimum operating) voltage designed by the manufacturer and certified by the listing authority to work!

e.g.

24 volts nominal primary supply – 20 volts minimum required supply = 4 volts maximum losses (16.6%)

And..

20 volts nominal secondary supply – 16 volts minimum operating supply = 4 volts maximum losses (20%)

Then Res Ipsa Loquitur: (The Thing Speaks For Itself)

Maximum Voltage drop = starting voltage - minimum operating voltage

It is only when the MINIMUM OPERATING VOLTAGE designed by the manufacturer (which can be ANY number smaller than .85 times the nameplate) is subtracted from the NAME PLATE VOLTAGE designed by the manufacturer that the maximum voltage drop is defined. And brother there aint no percentage in arguing that.

Can we voltage drop the subject now?

* footnote

When the minimum performance is defined as “capable of performing its intended functions at 85 percent and at 110 percent of the nameplate voltage”;

(NFPA1996 art. 1-5.5.1) it sets a minimum and maximum parameter. 85 percent is the maximum minimum parameter and 110 percent is the minimum maximum parameter

Letter from AHJ

Hello xxxxx,


Below is typical language used in plan review letters regarding vo

drop calculations. This is the current policy enforced by the Dal

Fire-Rescue Department for all fire alarm plan reviews.

"All circuit voltage drops shall be limited to no more than 15%, o

volts for a 24 volt system. Circuit #1 serving the 2nd and 3rd fl

of Building Type I exceeds the maximum allowable voltage drop (see

attached calculation sheet). The contractor shall provide circuit

that the maximum voltage drop does not exceed 3.6 volts. Provide

battery capacity and voltage drop calculations that reflect change

the alarm notification system."

As mentioned, a copy of the Dallas amendments to Chapter 9 of the

International Fire Code are available from myself or any of the ot

engineers at 320 E. Jefferson St., Room 105. A complete set of th

Building and Fire Code amendments can be purchased at the cashier'

in Room 118.

Please let me know if you have any questions or require additional

information.

Regards,

xxxxxxxxxxxxxx

Senior Fire Protection Engineer

Development Services Department

xxxxxxxxxxxxxxxxx

Dallas, TX

Appliance Lump Sum, End Evaluation Method

NFPA 72 Handbook Equation
(1999)
• Appliance lump sum, end load evaluation:
Vload =Vterm − (Iload )(Rconductors)
Where:
Vload = minimum operating voltage of the appliance
Vterm = 20.4 (unless specified by mfr.)
Iload = Total Current draw of the connected appliances
Rconductors = total conductor resistance

Designing Conduit Runs EIA/ TIA vs. NEC

Before
we start to discuss the differences between the EIA/TIA-569 (Commercial
Building Standard for Telecommunications Pathways and Spaces) and the
National Electrical Code (herein referred to as “NEC” or “the code”),
as they pertain to conduit installations, let’s keep in mind, the fact
that the NEC is the law of the land, and the EIA/TIA-569 is a recommended standard.

Important Note: Always refer to local codes, in addition to the NEC, or 569, when planning, or performing an installation.

If
you ever read through the NEC, you will notice that along with almost
every rule, there will be a list of exceptions to the rule. The
exceptions recognize the fact that these installations are done in the
real world. There are many instances where a rule may not apply, or not
be practical, or may make certain installations impossible to do. The
writers of the code, on a regular basis, receive feedback from, and
consult with, many building industry sources, and make changes
accordingly. The NEC is revised every three to four years.

It
seems that the 569 standard is much more stringent on many aspects of
conduit installations, than the NEC. There can be many circumstances
where a telecommunications professional, with good intentions, may have
no other choice but not to comply with the 569 standard. I do not
believe that any law, set of rules, guidelines, etc., that are made so
strict that people choose (or have no other choice) not to follow them,
will properly serve the group of people (or industry) that they were
designed for. Luckily, the TIA/EIA-569 standard is a living document,
which means that it can be revised or changed, due to many factors,
including feedback from industry professionals. At the end of this
article, we will list, ways in which you can voice your opinion to the
TIA/EIA committees.

Sizing Conduits - Fill Factor

When
designing a conduit run, the most important decision that you will have
to make is it’s size. Consider not only the cables that will be
installed now, but the likelihood of having to add cables in the
future. “Fill factor” or conduit fill, states the maximum
amount of space that the installed cables should occupy in a given size
conduit, expressed as a percentage of the interior volume. On the
subject of fill factor, the NEC and EIA/TIA-569, are for the most part,
in agreement (more about the comparison later). Consider the
percentages that are mandated by the “1996 National Electrical Code”,
Chapter 9, Table 1 (see table). Note that where the table indicates
“number of conductors”, the word “conductors”, may also indicate
“multi-conductor cables”. For example, if we were planning to pull
three cables into a conduit, the combined cross-sectional area of the
three cables, must not exceed 40% of the conduit’s interior volume.
Because the table does not specify high or low-voltage cable(s), it can
apply to both.


Number of Conductors 1 2 over 2
Percent Fill(all conductor types) 53% 31% 40%

Table 1

Table 1

The fine print note that follows, states:

“Table
1 is based on common conditions of proper cabling and alignment of
conductors, where the length of the pull and number of bends are within
reasonable limits. It should be recognized that, for certain
conditions, a larger size conduit, or lesser conduit fill should be
considered.”


My interpretation of the fine print note is
that a certain amount of overkill in conduit sizing is OK, and
encouraged! Notice that the smallest percent fill, on the table is for
two conductors (no more than 31% of the conduit’s interior volume). The
reason for this is that two conductors, of the same size, collectively
form an oval shape. One, or any number of cables greater than two, will
tend to form a circular shape (see figure ZZ).

Calculating Fill Factor

When
we calculate fill factor, normally we start out knowing how many cables
we need to get from point A to point B. The question is: What size
conduit? Or, sometimes a conduit will be existing and we will need to
calculate the amount of cables that we may install in it. Once you
master the principles of NEC Table 4, and the formula for
cross-sectional area, you can easily find either.


Electrical Metallic Tubing
Trade Size
Inches
Internal Diameter
Inches
Total Area 100%
Sq. In.
2 Wires
31%
Sq. In.
Over 2 Wires 40%
Sq. In.
1 Wire
53%
Sq. In.
½
¾
1


2

3

4
0.622
0.824
1.049
1.380
1.610
2.067
2.731
3.356
3.834
4.334
0.304
0.533
0.864
1.496
2.036
3.356
5.858
8.846
11.545
14.753
0.094
0.165
0.268
0.464
0.631
1.040
1.816
2.742
3.579
4.573
0.122
0.213
0.346
0.598
0.814
1.342
2.343
3.538
4.618
5.901
0.161
0.283
0.458
0.793
1.079
1.778
3.105
4.688
6.119
7.819

Note: Shown is only the Electrical Metallic Tubing (EMT) portion of NEC table 4.


“The Table”

Illustration
XX is the NEC table 4, which expresses the exact internal diameters of
each of the conduit types, in the various sizes. The diameters are then
converted to total area (similar to the way we would calculate the
square footage of a room, but using a formula for circular area).The
percentages (31, 40, and 53 percent) are then calculated for us.
Contrary to popular belief, different types of conduit have slightly
different interior diameters. Note that in table 4, there are separate
charts for each of the conduit types.

For example, three or more
cables, installed in a 2” EMT conduit, should not have a combined
“cross-sectional area” of more than1.342 square inches. This figure was
based on the fact that a 2” EMT has an internal diameter of 2.067
inches, the total area (in square inches) is 3.356 inches, 40% of 3.356
is 1.342.


“The Formula”

To find out how much area a cable (or cables) will take up, we can use the following formula:

Cable Diameter Squared x 0.7854 x = Cross Sectional Area

Note: The number 0.7854 is arrived at by dividing Pi by four (3.1416 Div. 4 = 0.7854)

Once
you have determined the “cross-sectional area”(CSA), for each cable,
simply add the CSA (or multiply for same size cables) for each cable to
find the Total CSA.

We will now try out the formula and the
table to see what size EMT conduit will be required for sixteen (16)
category-5 cables. For the purpose of this illustration, assume that a
typical cat-5 cable has an outside diameter of .24 inches (just under a
quarter inch). First we will find the cross-sectional area for one
cable, and then multiply by sixteen, for the total CSA, of the cable
group.

Calculate:

(0.24 Squared)

(.24 x .24 = 0.0576)

0.0576 x 0.7854 = 0.045239 square inches per cable in Cross Sectional Area.

0.045239 x 16 = 0.723824 of total square inches in Cross Sectional Area.

Now
that we have determined how much area, sixteen category-5 cables will
occupy, we can consult Table 4, the section for “Electrical Metallic
Tubing”, under the column “over 2 wires - 40%”. As we look down the
column, we must look for a number that is equal to or greater than our
number. The smallest number that fits the bill is 0.814. Staying on
that horizontal line, looking to the left, we can see that the minimum
size EMT conduit, for sixteen cat-5 cables would be 1 ½” trade size.

569 Compliance

Don’t
be so fast to pull those 16 cables in the 1 ½” conduit. If this is to
be a 569 compliant installation, we must abide by 568: 4.4.2.4 “Any
single run, extending from a telecommunications closet shall serve no
more than three outlets.” I checked with the TIA, their definition of
an outlet, boils down to however many jacks, or couplers, can fit on a
wallplate. So an outlet can translate to a single, or six or more
cables. So if the installation required a single cat-5 to a wallplate,
we would only be allowed to pull three cables into the 1 ½” conduit or
a 4” conduit for that matter. In this scenario, we would actually use a
¾” conduit Table 4.4-1 indicates that a ¾” could fit three (3) .24
diameter cables. Apparently, this chart was based on the NEC
guidelines, but only takes into account EMT conduit, even though that
fact is not stated.

It is interesting to note that in the city
of Chicago, conduit is required, for low voltage installations in the
ceilings, and walls of all hi-rise buildings. Imagine an installation
in Chicago that needed to be 569 compliant. If a closet served two
hundred outlets, there would have to be sixty-seven (67) conduits
extending from this closet.

“Boxes”

568: 4.4.2.6.3 states
“Boxes shall be placed in a straight section of conduit and not used in
lieu of a bend.” The corresponding conduit ends should be aligned with
each other. Refer to the diagram AA, Figures A, B & C, are three,
traditional methods of horizontal conduit distribution, none are 568
compliant. Figure D, is compliant, however it would only be practical
for use as a ceiling pull box, Figure E , would be the only compliant
way to run conduits for horizontal distribution, and be able to have
one homerun conduit serving three outlets.

Conduit Bends

There
are two aspects of conduit bends for discussion, number (of bends) in
one run, and radius. As we will see the 569 standard is much more
stringent on both of these aspects, than the NEC.

Number in One Run

NEC
346-11: “There shall be no more than the equivalent of four quarter
bends (360 degrees total) between pull points, e.g., conduit bodies and
boxes.”

TIA/EIA-569, 4.4.2.1: “No section of conduit shall be
longer than 30m (100ft) or contain more than two 90 degree bends
between pull points or pull boxes.

Bear in mind that the NEC
rule applies to high voltage circuits as well as low voltage. High
voltage can be 10, 25, or 50KV (10,000 volts, 25,000 volts and 50,000
volts, respectively). If these highly sensitive, high voltage cables,
can be safely installed by following 346-11, then why can’t low voltage
cables? Why put a distance restriction? Communication cable may be
sensitive, but it is not quite a string of egg shells. It would be
interesting to know if the 569 committee did any studies on the
stresses, and their effects, that a communication cable will undergo
when pulled into conduits, under various conditions. Were the
guidelines based on hard facts, or just the rule of thumb that a major
amount of overkill will surely get the job done?

Radius of Bends

To
comply with the NEC, you need only use a standard conduit bending
devise, or purchase pre-fabricated bends. Manufacturers of conduits and
their bending apparatuses, make their products to comply with the NEC
table 346-10. The NEC 346-10, has only two categories: Conductors
without lead sheath, and, Conductors with lead sheath. Lead sheathed
cables, once commonly used for electrical service entrances, have all
but been replaced by newer insulation’s. For our purposes, we will only
refer to the column for Conductors without lead sheath.

On bend radius, the TIA/EIA-569, 4.4.2.2, states the following:

“The
inside radius of a bend in a conduit shall be at least 6 times the
internal diameter. When the conduit size is greater than 50mm (2
inches), the inside radius shall be at least 10 times the internal
diameter of the conduit. For fiber optic cable, the inside radius of a
bend shall always be at least 10 times the internal diameter of the
conduit”


To compare the NEC’s standards to those of the
TIA/EIA-569, I have compiled the following chart. The chart has three
columns. The numbers in the columns indicate radiuses, expressed in
inches. Since standard conduit bends comply with the NEC, the first
column, labeled “standard” refers to NEC guidelines. Column 2,
indicates 6 times the conduit’s internal diameter, as recommended by
the 569 standard for communication cables (except fiber optics) pulled
in conduits 2” or less. Column 3 indicates ten times the conduit’s
internal diameter, as mandated by 569, for fiber optics, and conduits
over 2”.


EMT Conduit Size Standard 6 x Diameter 10 x Diameter
½” 4” 3.73” 6.22”
¾” 4.5” 4.94” 8.24”
1” 5.75” 6.29” 10.49”
1 ¼” 7.25” 8.28” 13.8”
1 ½” 8.25” 9.66” 16.1”
2” 9.5” 12.4” 20.67”
2 ½” 10.5” N/A 27.31”
3” 13” N/A 33.56
3 ½” 15” N/A 38.34”
4” 16” N/A 43.34”

Notes:

1.
The chart above is based on the internal diameters of EMT conduit.
Fractions of an inch are expressed in a decimal format, rounded to the
nearest hundredth of an inch.

2. Numbers in red are higher than the standard radii, required by NEC 346-10 .

As
we can see by the chart above, the only standard bend that complies
with the “six times” rule is for ½” EMT. All other conduit sizes would
require custom bending. In addition, most of the radiuses listed in the
“ten times” column, would be cumbersome to handle and install, to say
the least.

In Conclusion

As a medium, conduit is
a viable and practical distribution system for many applications.
Unfortunately, the 569 standard, all but eliminates the words “viable
and practical”. I think most would agree that some changes need to be
made to the 569 standard, pertaining to conduit installations. If you
use the information presented here along with some good common sense,
you should be able to attain a trouble free and professional
installation.
Reprinted by permission from www.LANshack.com

The Polar and Rectangle Array commands in AutoCAD

Thes are powerful and fun commands in AutoCAD, the possibilities are endles...

Sleeping Rooms

A visual appliance used to awaken a sleeping, hearing impaired person shall be permanently installed and located 16 feet maximum from the pillow or “head of the bed” location, measured horizontally. Wall mounted appliances located at least 24 inches below the ceiling shall have a minimum effective intensity of 110 candela. Ceiling mounted appliances and wall mounted appliances located less than 24 inches below the ceiling are to have a minimum effective intensity of 177 candela. If a suite contains more than one sleeping area, an appliance shall be provided in each sleeping area.

Audible Notification

Audible notification appliances intended for operation in the public mode shall have a sound level of not less than 15 dBA above the average ambient sound level or 5 dBA above the maximum sound level having a duration of at least 60 seconds, whichever is greater, measured 5 ft. above the floor.

Mounting Height

Wall mounted appliances are to be located 80 inches minimum and 96 inches maximum above the finished floor. NFPA clarifies this by requiring the entire lens to be within 80” to 96” above the finished floor.

Synchronization

Synchronized Strobes


ADA recommends that a composite flash rate in excess of 5 Hz shall be avoided for multiple strobes installed in the field of view, and indicates that the use of synchronized strobes should provide an acceptable alternative when more than two strobes are installed in the same field of view. In addition, it is important to remember that in order to meet ADA’s concern for effectively alerting the hearing impaired, strobe flash rates must be a minimum of one flash per second, across the Regulated* voltage range. Using lower intensity synchronized strobes can provide better alerting since occupants can often directly view the strobe flash. In addition, the selection of strobe intensity and location is simplified.

Synchronized Temporal Pattern

NFPA 72 (Chapter 6, Sections 6.8.6.4) requires the use of the temporal pattern (Code 3) audible signal for public mode evacuation. To insure that the sounds from multiple signals within a specified area do not overlap and interfere with the distinctive temporal pattern, the devices shall be synchronized. Since a majority of fire alarm installations are retrofit applications, the ability to install synchronized appliances using two wires becomes a significant issue. With the use of two (2) wire signaling appliances, such as The Wheelock Series AS and NS, existing wiring may not have to be replaced unless the current capacity of the existing NAC circuit needs to be increased.

Evacuation Signal Zoning

NFPA 72 (Chapter 6, Section 6.8.6.4.3) requires synchronization of notification appliance circuits within an evacuation signaling zone.

Clarification on Notification Issues by Mike Ryan

Not to put too fine a point on it but rather to be precise;

Wall mounted (visible) appliances are to be located (between) 80 inches minimum and 96 inches maximum above the finished floor. (i.e. 80" AFF to the bottom of the lens so that the entire lens is no lower than 80" and no higher than 96" AFF to the top of the lens, so that the entire lens is no higher than 96")

Now if you want to pick a fight consider this;

The mounting criteria for Visible appliances differs significantly from wall mounted Audible appliances, which have a minimum mounting of 90" AFF as measured to the top of the appliance.

You will note, when Combination appliances are mounted so as to satisfy the Audible criteria, (90" AFF top) the appliance remains within the range allowed for both Audible AND visible appliances, but not the other way round.

In other words, where combination appliances (Audible & Visible housed in the same enclosure) are mounted at 80" AFF to the bottom (of the lens) the resulting installation will make the audible too low.

The reasoning stems from the prohibition against mounting an Audible Appliance so low that the resulting SPL would exceed 120dBA at the minimum hearing distance

Mike Ryan - Alarm Express

Old Coded Manual Pull Boxes


Frontline Fire Protection, Inc.




17815 Davenport
Suite 180
Dallas, Texas 75252
972-818-9565

AutoCad Tip o fthe Week #1 - Binding

External references are overused and are a nightmare. They make
someones job easier, but mine stupid hard.
Bind all XREF's, first make sure
that they are not locked on the layer, then goto
View-XREF manager, or the like, and bind everything you can.
The paths to missing files are also visible in the dialog box.
Then, feel free to explode, then turn off all the junk in
layer control, then there is a much easier drawing to work with.
Don't forget to purge.