Lesson 12

Measures and payback

The audit of lessons 10 and 11 found what the plant wastes. This lesson turns the findings into a plan: a copy of the network as it will be, the measures made in it one by one, the two versions side by side — and, from the difference, what each measure pays back.

Scenarios: one file, several networks

A scheme file can hold several versions of the same network. The first one is the baseline — the plant as it is. The others are copies you change: the plant as it will be. A measure is never applied to the baseline, so the audit of the real plant always stays in the file.

So the comparison is between networks under the same tariffs and hours.

1. Start from the audit of lesson 11

   Open lesson11.gf9, save it at once as lesson12.gf9 and press Audit — the measures are searched only by the button. Check COMP against lesson 11:

   Row, COMP	Expected	Unit
   Audit pressure	6.5	bar
   Audit power	8.7	kW
   Qn average	995.68	NL/min
   Min audit pressure	6.1832	bar
   Power calibration	0.752
   Annual cost calibrated	2172	€

   PG reads 6.35 bar, FM 1966 NL/min: both within accuracy. If PG still reads 6.0 from the experiment of lesson 11, set it back first.

2. Read the measures at the measured energy

   Open the Problems panel and click the green Recommendations counter. The measures of lesson 10 are there again, with the leaks of lesson 11 — but every figure in euros is now multiplied by Power calibration 0.752: the money is counted at the energy the meter measured, not at the price of air of Input parameters.

   Problem	On	Value
   Leakage is above the target: a repair programme pays back (25 % of the average demand, 249.3 NL/min; target 5 %)	COMP	458 €/yr
   The compressor setting can be lowered (6.5 bar → 6.183 bar without a pressure deficit at maximum load)	COMP	154 €/yr
   [-] 2 × TUBE_BLOWING — the consumer gets more pressure than it needs and uses more air	B2, B1	209 €/yr
   6.462 bar instead of 4 bar; 188.6 NL/min instead of 126.4 at average load	B2	136 €/yr
   6.45 bar instead of 6 bar; 557.8 NL/min instead of 524.1 at average load	B1	73 €/yr
   Undersized pipe: high air velocity and pressure loss (9.56 m/s and 414 Pa/m at maximum load; limits 9 m/s and 200 Pa/m)	T1	0.015 bar
   The compressor runs at a low average load (average 39.8 % and peak 78.4 % of the capacity 2500 NL/min)	COMP

   The rows name the elements by ID and card label: COMP is ID7 compressor, B1 and B2 are ID1 and ID2, T1 is ID3 D-1 IN_EQUAL TEE.

   Select a row and look at the Apply button on the bottom bar. It is enabled only for a measure that knows exactly what to change in the scheme:

   Can be applied	What Apply changes
   The compressor setting can be lowered	Audit pressure := Min audit pressure, rounded up to 0.01 in the unit of the row
   Undersized pipe, This pipe limits the setting	the pipe (or tee) is replaced by the catalog size the measure names — only when that size meets the limits
   Condensate drain loses compressed air	the drain becomes an electronic zero-loss one

   Select Undersized pipe of T1: Apply is enabled, and the detail panel names the tee it would put in — d-1.5 in_Equal Tee (37.5 mm), 3.6 m/s and 44 Pa/m. Leave it for now; the last item of Try this next applies it. A regulator before a consumer, a leak repair, another compressor control — these you make by hand in a scenario, as the next steps do.

3. Make a copy of the plant for the plan

   Choose Scenario → New from current.... The program offers the name Proposed; press OK. The copy becomes the current scenario:

   • a strip above the sheet reads SCENARIO: Proposed (the baseline reads BASELINE: As is, in orange); a click on it lists the scenarios;
   • the window title ends with — Proposed.

   Why not press Apply on the lower setting straight away? Pressed in the baseline, Apply makes this copy by itself. But the lowest setting depends on the network: B2 takes 6.46 bar where 4 would do, and every litre it blows runs through P1 and takes pressure from B1. Put the regulator in first, and the setting can go lower. Make the measures that change the network first, the setting last.

4. Put a regulator before B2

   The regulator before B2 is a measure the program cannot apply: it means a new element on the sheet. Make sure you are in Proposed, then do what lesson 8 did: release the connection P3 → B2, place a regulator from Regulators — MD1-R00-1/8 — between them, join both sides and rename it R2 (F2). Set:

   Item	Row	Value
   R2	Pout	4 bar	from the list; what Audit required pressure of B2 asks for

   Pout of a catalog regulator offers only the set points at which the maker measured its flow curves — the list of the card. The knob of a real regulator turns smoothly; for a setting between the curves there is Audit set point: any number, computed on the nearest curve. Leave it empty for now.

   Press Audit and click R2. A regulator does not give exactly its set point: its outlet droops with the flow along the curve of the card, and this one also needs at least 1 bar across it (dPmin).

   Row, R2	Value	Unit
   m	0.0083	kg/s
   Pin	6.3765	bar	well above 4 + 1
   Pout actual	3.9786	bar	0.02 bar below the set point: the droop
   delta P	2.3979	bar	the pressure the regulator takes away on purpose
   Qn	419.374	NL/min	what B2 now blows while it runs
   Opening	33.5935	%
   Required set point	4.0214	bar	what gives B2 its 4 bar

   A 1/8 in regulator at a third of its opening droops very little — but enough: B2 gets 3.9786 bar, 0.0214 short of its required 4, and shows Pressure deficit 0.0214 bar. While a consumer is short, no setting of the compressor cures it — the regulator holds its own outlet — so the list has no The compressor setting can be lowered now. Required set point is the answer: keep Pout at 4, type 4.03 into Audit set point of R2 — the estimate rounded up — and press Audit again. The droop of the curve moves with the set point:

   Row	Value	Unit
   R2, Pout actual	4.006	bar	what B2 gets
   R2, Set point vs design	0.03	bar	Audit set point − Pout of the card
   B2, Pressure margin	0.0062	bar	the deficit is gone
   B2, Qn artificial demand	0.75	NL/min	what the 6 mbar of margin costs

   B2 blows about 420 NL/min while it runs instead of about 610. Less air runs through P1 at maximum load, B1 gets more pressure, and The compressor setting can be lowered is back in the list — now from 6.5 bar to about 6.16 bar, lower than the 6.183 of the baseline.

5. Apply the lower setting

   Select The compressor setting can be lowered and press Apply. Proposed is not the baseline, so the change goes into it: Audit pressure of COMP becomes 6.17 bar — the minimum rounded up, so that the setting never ends below the one found. The program runs the audit again and reopens the Problems panel with a fresh list; the measure is gone from it, and the change is written into the journal of measures of Proposed — step 7 shows where to read it.

   Row, COMP	As is	Proposed	Unit
   Audit pressure	6.5	6.17	bar
   Qn	1960.35	1708.22	NL/min
   Load at maximum	78.41	68.33	%
   Qn leak total	247.39	236.904	NL/min
   Qn average	995.68	899.037	NL/min
   Cs	2889	2528	€
   Power calibration	0.752	0.752
   Annual cost calibrated	2172	1900	€
   Leak share	25.04	26.52	%
   Leak cost	544	504	€
   Min audit pressure	6.1832	6.1613	bar

   Power calibration in Proposed is not measured again: the meter measured the plant as it is. Its note reads From the baseline scenario — every scenario is priced at the energy the baseline measured, so the scenarios can be compared at all.

   Check by hand: the cost of the plan

   Without a machine the cost is the price of air corrected for the pressure: 0.024 €/Nm³ × (p / 6)^0.603 × Qn average × 1920 h, then × 0.752. As is: (6.5 / 6)^0.603 = 1.0495, 995.68 NL/min → 2889 € → 2172 €. Proposed: (6.17 / 6)^0.603 = 1.0170, 899.04 NL/min → 2528 € → 1900 €. The regulator and the lower setting together save 272 € a year.

   PG and FM now carry warning signs. Their readings, 6.35 bar and 1966 NL/min, were taken on the plant as it is; in Proposed the setting is lower and B2 takes less, so the model no longer matches them: PG shows Pressure deviation 0.28 bar (the model gives 6.07), FM Qn deviation about 258 NL/min. A reading belongs to the network it was taken on. Leave the two rows as they are: the comparison counts them as two warnings of Proposed, and step 7 shows where. Clearing Audit measured of PG and FM in Proposed removes the signs — the baseline keeps its readings either way.

6. Model the leak repair in a scenario of its own

   The biggest measure is the leaks, and it has two prices: a survey and the repairs. Keep it apart so that its payback can be seen on its own. With Proposed current — the regulator in place, the setting at 6.17 — choose Scenario → New from current... and name the copy Leaks repaired.

   A repair cannot be applied by the program: you tell the audit what the leak test will show after it. Say the survey brings the leaks down to a third. On COMP:

   Row	As is	Leaks repaired
   Audit leak t load	6 s	2 s	loaded
   Audit leak t unload	54 s	58 s	unloaded
   Audit leak test pressure	6.5 bar	6.5 bar	copied — leave it

   Press Audit. The test gives 2500 × 2 / 60 = 83.3 NL/min at 6.5 bar; at the lower network pressure of 6.17 bar the leaks are a little less, 79.03 NL/min — a third of the 236.9 of Proposed.

   Row, COMP	Expected	Unit
   Qn leak at test	83.33	NL/min
   Qn leak total	79.03	NL/min
   Qn average	740.6	NL/min
   Leak share	10.74	%
   Annual cost calibrated	1565	€

   The leak share, 10.74 %, is still twice the 5 % target of the measure Leakage above the target, which stays in the list of Leaks repaired — a third of the leaks is still a lot for a network this small. The target would take a test of under 1 s loaded in a minute.

7. Compare the scenarios

   Choose Scenario → Compare scenarios.... The Figures tab has a column per scenario and the differences against the baseline. The second header row shows the state of each column: audit — current; outdated — its input has changed since; press Calculate all to bring every column up to date.

   Row	As is	Proposed	Leaks repaired	Unit
   Source pressure	6.5	6.17	6.17	bar
   Source pressure, average load	6.5	6.17	6.17	bar
   Lowest setting without deficits	6.183	6.161	6.15	bar
   Demand at maximum load	1960	1708	1552	NL/min
   Demand at average load	995.7	899	740.6	NL/min
   Compressor load at maximum	78.41	68.33	62.08	%
   Leakage	247.4	236.9	79.03	NL/min
   Leakage share	25.04	26.52	10.74	%
   Model calibration by Audit power	0.7516	0.7516	0.7516
   Annual electricity, calibrated	16.7	14.62	12.04	MWh
   Annual electricity cost, calibrated	2172	1900	1565	€
   Cost of compressed air (estimate)	2889	2528	2082	€
   CO2 (estimate)	8889275	7778139	6407197	g
   Consumers short of pressure	0	0	0
   Errors and warnings	0	2	2
   Saving potential of the audit measures	715.3	451.5	118.4	€

   The share of leaks in Proposed is higher than as is, although the leaks themselves are a little lower: the consumers take less, so the same leaks are a bigger part of what is left. Read the absolute rows, not only the shares. The two warnings of Proposed and of Leaks repaired are PG and FM with the readings of the baseline (step 5). The leak share is not among them: the warning thresholds of the leak share count only leaks drawn on the sheet, leak orifices; the leaks found by the test give the measure Leakage above the target instead.

   The last row, Saving potential of the audit measures, is what the measures the audit still finds in each scenario would save: 715.3 € as is, 451.5 € after the regulator and the lower setting, 118.4 € after the repair too. What is left is mostly the rest of the leaks.

   Check that Leaks repaired was made from Proposed: its tab reads Leaks repaired (6) — the four changes of Proposed and the two leak test rows — and its demand at maximum load, 1552 NL/min, is below that of Proposed. A Leaks repaired made from As is, with 6.17 typed by hand, has no R2: its tab reads (3), B2 blows its 610 NL/min again, and at 6.17 bar B1 is short of pressure (When it does not work). A setting found with the regulator in place is only safe with the regulator in place.

   The tab Proposed (…) lists what differs from the baseline, element by element, with the measure that caused each change:

   Element	Change	Parameter	As is → Proposed	Measure
   COMP	Changed	Audit pressure	6.5 → 6.17	The compressor setting can be lowered
   R2	Added
   P3	Reconnected	Connected to	T1, B2 → T1, R2
   B2	Reconnected	Connected to	P3 → R2

   Four rows — hence Proposed (4) on the tab. On the screenshots the elements are named by their IDs, as in lessons 7–11.

   The regulator has no measure in the journal: it was placed by hand. The tab of Leaks repaired lists the same, plus the two leak test rows of COMP.

   The last tab, Measures, is the journal itself: every entry of every scenario in the order it was made, with what was changed and what the measure promised at that moment. Made by the steps of this lesson, it has two entries:

   Scenario	Element	Measure	What was changed	Air saved, NL/min	Power saved, kW	Saving a year, €
   Proposed	ID7 - compressor	The compressor setting can be lowered	Audit pressure: 6.5 -> 6.17 bar	36.5		141.8
   Leaks repaired	ID7 - compressor	The compressor setting can be lowered	Audit pressure: 6.5 -> 6.17 bar	36.5		141.8

   The element is named by its ID and card label. Leaks repaired has the same entry: a new scenario starts with a copy of the journal of the one it was made from. The regulator and the leak test are not there — the journal holds what Apply changed, cards replaced by hand and values set by the variant calculation, not every edit of a row. The euros are a promise: 141.8 € is what the lower setting was to save on the network with the regulator, at the moment it was applied. With two or more entries that promise a saving, a scenario ends with a bold Sum of the estimates; it is only a sum of promises, the result of them all is on Figures. Double-click an entry to go to its element in its scenario.

   The screenshots of this lesson show Measures (4): in their file the card of the flow meter had been picked from the catalog again in Proposed, so Proposed has one more entry, FM_1 — Card replaced by hand — Card: VD 500 -> VD 500, and Leaks repaired copied it with the rest (there it reads FM_1 - VD 500). The columns of power and money are further right; scroll the table to see them. That is harmless (see When it does not work); a card picked again promises nothing, so there is no Sum of the estimates either.

8. Work out the payback

   EasyCAS gives the saving of each scenario; the price of the work comes from your supplier. Divide one by the other. With the figures of the comparison and example prices:

   Measure	Saving a year	Investment (example)	Payback
   Lower setting, 6.5 → 6.17 bar	≈ 142 €	0 € — a setting	at once
   Regulator R2 before B2	≈ 130 €	120 € with fitting	≈ 11 months
   Leak survey and repair	335 €	900 €	2.7 years

   Proposed saves 2172 − 1900 = 272 € a year, the regulator and the setting together. The journal gives the share of the setting — the 141.8 € it promised on the network with the regulator; the rest, about 130 €, is the regulator. The leak repair is the difference between the next two columns, 1900 − 1565 = 335 €. Taken in this order the savings add up to the total, As is → Leaks repaired, 606 € a year (the change column of Leaks repaired: −606.3 €). Taken separately, each against the baseline, they would not add up: the regulator and the lower setting save part of the same air.

   The small plant, the small money. Two open pipes and 50 m of pipe on a 15 kW compressor working one shift give hundreds of euros, not thousands. The method is the same on a plant with forty consumers and a 160 kW room — and there the survey pays in weeks.

9. Save it

   Save lesson12.gf9. All three scenarios are kept in the one file, with the journal of measures; the comparison opens without recalculating. Lesson 13 splits the cost of the air between the departments of the plant.

When it does not work

Try this next

• In Proposed, set Audit set point of R2 to 3.8. B2 is short of pressure: the required pressure decides, not the measure's advice. Set it back to 4.03.
• In Compare scenarios, set Compare with to Proposed: the change columns now show what the leak repair adds on top of the setting and the regulator alone.
• Save the file as lesson12-try.gf9. In Proposed select Undersized pipe of T1 and press Apply: T1 becomes a d-1.5 in_Equal Tee. Measures now has two entries for Proposed and the row Sum of the estimates; Leaks repaired keeps its one — it was copied before. Go on with lesson 13 from lesson12.gf9.

Related help topics: