Standby mode has negligible consumption… but after two or three years?

The nine systems and their geography in my home

I cannot sleep. I lie on my back in the darkened bedroom, listening to my wife's slow, steady breathing beside me — the breathing of someone asleep without a single thought about lights, standby, kilowatts, or warm transformers. For her the world is simple: press the switch, light comes on; press again, light goes off. End of story. But for me, twenty-three minutes past midnight, after seeing that figure — 5.2 watts — simplicity has shattered like a mirror I have been gazing into, satisfied, for far too long. In the shards I see not the home I thought I understood but an electrical geography I knew almost nothing about.

I think of the nine systems. Their geography throughout the flat. The way I installed them, one by one, over the years, each with its own story, each bought for a particular reason, each promising something: economy, comfort, modernity. And now, in the dark, I revisit each of them, relive their history, try to understand how I arrived here — with 5.2 watts consuming permanently, drawn by lights that are supposed to be dead.

Hallway — Left and Hallway — Right. The first two. Installed four years ago, when I still harboured illusions about the smart home. I was in that phase of reading about the internet of things, automation, how the future means controlling everything from your phone. Our hallway is small — roughly four metres long, a metre and a half wide — a corridor between the front door and the rest of the flat. It had a single light: an old ceiling fitting with a sixty-watt incandescent, switched by a button near the door. Simple. Functional. But it seemed insufficient to me. It seemed to me that an entrance hallway needed ambient light, a warm sense of welcome when you came home. So I fitted two LED spots on the side walls — one left, one right — each connected to a small transformer hidden inside an electrical box in the wall. Modern, white, round spots that gave a soft, diffuse light. And controllable by WiFi, of course. Because why get up from the sofa to switch off the hallway light when you can do it from the app?

I remember fitting them. A Saturday afternoon, late March. My son was five and watched me drilling into the wall with the electric drill, fascinated by the noise and the dust. "Dad, are you making a hole in the house?" "Yes, but it's a good hole. For the light." "But we already had a light." "Yes, but this one will be better." He didn't understand why the new light was better than the old one, but he listened with the blind trust children have in their parents, assuming adults know what they are doing. Now, four years on, I wonder whether I knew what I was doing. Whether replacing a single simple bulb with two WiFi spots connected to permanently live transformers was really an improvement, or merely a needless complication, driven by my desire to appear modern.

Those transformers — two black plastic boxes, each about the size of a nine-volt battery — I mounted behind the wall, inside the electrical box. I never see them. But now, lying in bed, I know they are there. I know that each one consumes something — probably half a watt, maybe more, given their age and likely degradation — and that their combined draw contributes to the 5.2 watts I saw on the meter. And I can do nothing. Because to switch them off completely I would have to cut the power at the distribution board. And if I cut the power, I lose WiFi control. And if I lose WiFi control, then why did I install them in the first place?

It is a perfect vicious circle. I fitted the lights to have control. But to have control, they must stay permanently live. And if they are permanently live, they permanently consume. So control has a price. Always. In every second, every minute, every hour of every day since the day I put them up.

The calculation: four years × 365 days × 24 hours = 35,040 hours of continuous consumption. If each transformer draws half a watt, the two hallway lights together have used: 35,040 hours × 1 watt = 35 kilowatt-hours. At the average tariff over four years — perhaps sixty bani per kilowatt-hour — that is roughly twenty-one lei. Twenty-one lei to keep the hallway lights connected, available, waiting for my command from a phone, for all that time.

It is not a fortune. But something else unsettles me: I paid it without knowing. I paid a maintenance levy nobody explained when I bought the bulbs. The box said: "85% energy saving." It did not say: "Plus an ongoing connectivity charge: approximately five lei per year." That silence is what stings, here at midnight, whilst my wife sleeps peacefully and the hallway lights consume, in their stillness, the energy needed to stay awake.

Living Room — Centre. The main ceiling light. Four LED spots in a square white metal frame hanging above the oval table where we eat every evening. This is the showpiece. The system I invested most in — money, time, hope. I bought it three years ago, from a lighting shop in Iaşi, after comparing dozens of models online, reading reviews, consulting forums. I wanted something modern, something that looked good, something that could be controlled subtly — adjustable for the moment: strong light when we are eating, dim light for films, medium light when my son does his homework at the table. And I found the perfect model: four dimmable LED spots, WiFi-controlled, programmable scenes, compatible with Google Home and Alexa — neither of which I had, but I liked knowing I could.

It cost four hundred lei. A lot, my wife noted when I told her the price. But I argued: it'll last ten years, we'll save on electricity, it'll be far more comfortable than the old one, which only had one brightness and made an irritating sound when it warmed up. And the new fitting is magnificent. At full brightness, the light is white, pure, distributed evenly across the table. At minimum it is warm and intimate, like candlelight. I can adjust it from the app, sliding from cool to warm, from bright to dim. It is total control. It is the satisfaction that technology gives you when it works exactly as it should.

But its transformer — a black box, larger than the hallway ones, roughly the size of a packet of cigarettes — is mounted in the ceiling above the metal frame, and it consumes. I know it consumes now. Because last night, before the experiment, I touched it and it was warm. Not hot enough to burn my fingers — warm, just. The temperature of a living body. And living bodies consume energy to remain alive.

Bedroom — Left Lamp and Right Lamp. Two bedside lamps, one on each nightstand. Bought two years ago, when we renovated the bedroom and decided we wanted "perfect symmetry" — my wife's phrase, not mine, but I agreed. Two identical nightstands in solid oak, with a lamp on each. Modern, metallic, with a cylindrical fabric shade and an integrated LED — and WiFi control, of course. Because why reach out and press a button when you can say "Hey Google, turn off the bedroom light"? We didn't have Google Home then, but I bought one shortly after the lamps, precisely to justify the voice-control feature.

I remember the evening we fitted them. Autumn, November, rain outside. Our son was already asleep. My wife and I unpacked the lamps, positioned them on the nightstands, plugged them in, configured them in the app. Then we switched off the main light and turned on only the lamps. Warm, diffuse light spreading across the walls in a yellow-orange halo. My wife smiled. "It's perfect. Exactly what I wanted." And we sat there in the lamplight, each reading — me something of Jung's on archetypes, her a novel about a woman fleeing a violent husband and rebuilding her life in a small coastal town. We felt at ease. We felt we had created an intimate, warm, modern space.

Now, two years on, I know that we created a space that permanently consumes. Because these lamps — though they have integrated LEDs, though they are "efficient" — also have transformers. Small ones, hidden in the metal base, but transformers. And transformers consume. Always. Even now, with the lamps off and both of us lying here — or me lying here, unable to sleep — the transformers are awake, waiting for the voice command or the app command that will not come until morning.

Kitchen — LED Strip. My favourite system. Or it was, before I started thinking about consumption. A strip of LEDs mounted beneath the kitchen cupboards, along the worktop. Three metres long, fifty-four individual LEDs, each the size of a coffee bean, each capable of emitting white or warm light depending on the setting. It is perfect for cooking: when I'm chopping vegetables, when I'm preparing dinner, when I'm making coffee in the morning — the light comes from exactly the right angle, top-down, illuminating the worktop without casting shadows. I fitted it two years ago, during the same renovation period, and was very pleased with myself. An elegant, practical, modern solution. Until I started thinking about the transformer.

The LED strip's transformer is larger than the others — the size of a pack of playing cards. Mounted in the cupboard under the sink, connected to the socket, feeding the whole strip through a thin cable that runs up under the furniture and connects to the end of the strip. This transformer has one specific detail: a small blue LED that stays on at all times, even when the strip is off. A "standby indicator LED" — to show you that the device is connected and functioning. When I fitted it, that blue LED seemed useful. Now it seems like a taunt. A small blue eye watching me from the cupboard under the sink, saying: I never sleep. I always consume.

Bathroom — Mirror and Bathroom — Shower. Three spots in total: two above the mirror, one above the shower cubicle. Installed three years ago when we renovated the bathroom. Small round silver recessed spots, set into the ceiling. They give strong white light, perfect for shaving, for makeup, for all the rituals of morning and evening. Controlled separately: I can switch on just the mirror spots for teeth-brushing, or just the shower spot for bathing, or all three for maximum light. Total flexibility. Total control. And, naturally, total consumption.

The transformers for the three spots are mounted above the false ceiling, in the narrow void between the real ceiling and the plasterboard. I have never seen them since they were installed. But I know they are there. Three black boxes, each connected to its spot, each consuming. In every second. In every minute. In every hour.

I imagine that space between the ceilings. I imagine the three transformers huddled together, warm in the dark, vibrating faintly as current passes through them, waiting for the command that will come tomorrow morning when I walk in to brush my teeth. Waiting with infinite patience, mechanical devotion, drawing their quiet maintenance tax for the privilege of being ready when I need them.

And I realise something: I have turned this flat into an organism that never sleeps. An organism whose vital organs — transformers, circuits, capacitors — function continuously, maintaining the state of wakefulness necessary for me, their master, to command instantly whenever I desire. I have created a mutual dependency: I depend on them for comfort; they depend on me for existence. And in exchange for this dependency, I pay constant tribute. Five point two watts. Thirty lei a year. And probably more, once I have measured a full twenty-four hours.

I wonder: if I had known, eight years ago, when I bought my first smart LED bulb — if I had known then everything I know now about standby, about transformers, about phantom consumption, about the hidden costs of modern comfort — would I still have bought all of it? Would I have installed all of it? Would I have turned my home into an electrical organism that never sleeps?

I don't know. And that troubles me more than anything else: that even with all my knowledge, even after the experiment, even after hours of calculations and analysis — I still do not know whether I would decide differently.

Perhaps that is the real discovery: not that the bulbs consume, but that I am willing to pay for a beautiful lie rather than accept an uncomfortable truth.

Waiting, and the temptation to look

I wake at six twenty-three, though I cannot recall having properly slept. For hours I floated in a strange borderland between sleep and wakefulness, in which figures drifted through the dark — 5.2, 5.2, 5.2 — repeating like an electrical mantra. My wife is still asleep, lying on her side with one hand under the pillow in that foetal position she always takes when she is truly relaxed. I envy her capacity to sleep without thinking about bulbs, transformers, phantom consumption. For her last night was like any other: in bed at eleven, twenty minutes of reading, asleep. Simple. Final. No obsessions.

I get up without noise. The parquet creaks under my weight — a creak I know so well I can anticipate precisely which board will emit which sound — but she doesn't wake. I go to the bathroom, wash my face with cold water, look in the mirror. Dark circles. Red eyes. Untidy hair. I look like a man who has stayed up all night calculating the consumption of light bulbs. Which, technically, I have.

I want to go to the kitchen. I want to check the meter. I want to see if the figure has changed, if consumption has risen or fallen, if the 5.2 watts has held constant or fluctuated. But I tell myself: not yet. The experiment must run twenty-four full hours. If I go now and check, I risk disturbing something, touching something, accidentally disconnecting something. I must resist the temptation. I must wait until midnight.

But how do you wait when the only thing in the world you are obsessed with is sitting in the kitchen, plugged in, measuring in silence a truth you do not yet fully know?

At seven, my son appears in the kitchen — dressed for school, hair standing in every direction — and watches me with curiosity. "Dad, did you finish the experiment?"

"The first experiment, yes. Today I'm doing the second."

"What are you measuring now?"

"Each light separately. To see which one uses the most."

He considers this, chewing a slice of bread and butter. "And if you find which one uses a lot, what will you do with it?"

I am still for a moment. What will I do, indeed? Throw it away? Replace it? Leave it consuming, carrying the guilt of knowing it is inefficient? "I don't know yet," I answer honestly. "First I have to find the culprit."

"Like a detective," he says, and his eyes light up. For him, this transforms my obsession into a story, a game. Dad the detective, hunting the bad bulb. It is a simplification that makes me smile, but which is also sad, because it reduces complexity to good and bad, guilty and innocent — when the reality is that all the bulbs are guilty, merely in different degrees, and the true guilty party is me, for installing them, knowing — or ought to have known — that nothing is free, that every convenience has a cost.

"Exactly like a detective," I confirm. "And when I find the culprit, I'll tell you."

IV. THE REVELATION — Reading the meter

1.3 kWh, or the collapse of certainties

Twenty-three fifty-eight. Two minutes left. I sit in the kitchen facing the digital meter as though facing an oracle about to reveal a truth I have suspected for months but lacked the courage to confirm until now. My wife is asleep. My son is asleep. The flat is asleep — or pretends to be, because I now know that within it pulses an electrical life that never ceases: an invisible breath consuming, always, in every second, whether I am awake or asleep, attentive or negligent.

I have checked the meter thirty-seven times throughout the day — yes, I counted, I noted each visit on a separate sheet, like a patient obsessively taking their temperature, hoping the fever has broken, knowing it has not.

Morning, at six: 0.041 kWh. At ten: 0.056 kWh. At noon: 0.067 kWh. At sixteen: 0.089 kWh. At twenty: 0.105 kWh. Now, at twenty-three fifty-eight: 0.128 kWh.

I look at the last figure. One hundred and twenty-eight milliwatt-hours. In nearly twenty-four hours. I do the calculation quickly, pencil on the margin of a sheet already covered in dozens of today's calculations — calculations crossed out, rewritten, redone, as though altering how I write the figures might change the reality they represent:

0.128 kWh in 24 hours = average consumption of 0.128 / 24 = 0.00533 kWh per hour = 5.33W continuously.

Five point three three. Almost precisely what I measured at the start. The consumption has been constant. The transformers have drawn steadily, without major fluctuation, 5.3 watts in every hour of every one of the twenty-four. They did not stop. They did not rest. They burned on slowly, like electronic candles, in the darkness of their plastic boxes set into walls and ceilings and cupboards.

But something does not add up. Something is profoundly wrong in the calculation, and I feel it as a nervousness rising from my stomach into my chest. Because I had assumed — calculated last night, verified dozens of times — that nine LED systems, each drawing roughly half a watt on standby, should consume roughly two to three watts in total, four at most if the transformers are old and degraded. But the meter shows five point three. Significantly more.

And then I understand something. Something I hadn't thought of last night, too focused on starting the experiment, too caught up in the need to confirm my suspicion. I realise I have made a fundamental error. I have forgotten something. That I have…

I rise from the chair so sharply that it scrapes the floor. I go to the meter. I look again. The screen shows two figures: 0.128 kWh — total energy consumed — and 5.4 W — instantaneous power. Five point four. Not five point three. It has risen by a tenth in the past three hours. Why? What is happening now, at midnight, that makes the consumption rise?

And then I see it. See what I should have seen last night but didn't, too absorbed in my obsession, too focused on the lights. I see the extension lead plugged into the meter. Six sockets. In five of them are the transformers of the nine LED systems — some transformers feed two or three bulbs through the same cable, which is why there are not nine separate transformers but only five. But in the sixth socket is connected… the router.

The router.

My stomach clenches. How is this possible? I unplugged it last night. I disconnected it at exactly midnight when I started the experiment. I watched its green LEDs go dark. I felt the house lose internet. It was dead. Completely dead.

But it was not. Because I had forgotten one detail. I had unplugged the router from its permanent socket in the hallway. But the previous evening, when I was preparing the experiment and connecting all the LED transformers into a single extension lead to measure them together, I had also connected the router into that same extension — because I wanted to measure the total standby consumption of the flat. Lights plus router. It made sense at the time. But then, when I saw the consumption was higher than expected, I completely forgot the router was included and assumed all five point three watts were coming from the lights.

A fool's error.

I crouch down and look at the extension lead. The router is there. Its green LEDs winking cheerfully, showing me it is on, functional, connected to the internet, and has been on all night, consuming. How much does a router consume? I know this. I had looked it up two weeks ago. A dual-band WiFi router draws between eight and fifteen watts when active, and between five and eight when in a power-saving mode — if it has one, which mine does not.

I sit back down. I take out my phone. I open the router app. I check the statistics. The router has been connected continuously for the past twenty-four hours. It has not entered any economy mode. It has been running at full capacity — probably because the devices in the flat, our phones, the laptops, my son's tablet, have been connecting to it periodically throughout the day. And it has been consuming.

I unplug the router from the extension lead. Its green LEDs go instantly dark. The flat loses internet for the second time in twenty-four hours. I watch the meter. The instantaneous reading falls: 5.4 W… 3.1 W… 2.9 W… 2.8 W… 2.7 W. It settles at 2.7 W.

Two point seven. That is the lights' consumption. The lights alone. Nine LED systems, all switched off through the app, all consuming together two point seven watts.

I do the calculation at once, pencil on the same sheet:

2.7W × 24 hours = 64.8 Wh = 0.065 kWh per day0.065 kWh × 365 days = 23.7 kWh per year23.7 kWh × RON 0.7 = RON 16.60 per year

Sixteen lei sixty. For the lights. Just the lights. The rest — the difference up to five point four — was the router. Which consumed 5.4 – 2.7 = 2.7 watts. Not eight. Not fifteen. Just two point seven. Exactly as much as the lights.

This doesn't make sense. The router's specifications say ten watts when active. But the meter showed only two point seven. Either the specifications are wrong, or the router uses far less when it is not transferring heavy data — at midnight, when everyone is asleep — or perhaps the specifications state the maximum worst-case draw, not the typical real-world consumption.

Whatever the reason, the result is: two point seven watts for the lights, two point seven watts for the router. Five point four in total. Which means approximately forty-seven kilowatt-hours per year. About thirty-three lei. For the lights and the router to stay awake while I sleep. For the house to be ready, instantly, the moment I need it. For the three seconds I would have to wait for the router to boot, for the five seconds for the transformers to warm up. For comfort. For convenience. For the illusion of instant control.

And the question becomes: is it worth it? Is thirty-three lei per year worth that?

At two minutes past midnight, after twenty-four hours of obsessive waiting, after a day in which I could concentrate on nothing else, after disrupting my family and unplugging the fridge and disconnecting everything — honestly, in this moment, I don't know.

But I know something. I know that I will never look at my lights the same way again. I will never press a button on the app without knowing that somewhere, in a plastic box inside a wall, a transformer is staying awake. I will never leave the router plugged in without realising it is consuming, always, even when unused. I will never blindly trust efficiency promises, specifications on boxes, marketing that tells me I have made an intelligent choice.

I will know. And that knowledge will change something inside me, even if it changes nothing in the flat.

The maths problem: where are those fifty watts?

I am sitting at the kitchen table with my sheet in front of me — the same sheet on which I have written dozens of calculations throughout the day, now filled with crossed-out figures, rewritten and recalculated, like a mathematical palimpsest documenting the evolution of my understanding, or more precisely, of my disappointment. Outside a dog is barking — probably the same shepherd from across the road, barking at cats, at cars, at the shadow of his own unease. The flat is silent. The fridge hums softly, reconnected, reclaiming its place in the household's electrical ecosystem. My wife and son are asleep. I ought to be asleep. But I cannot sleep. Because the figures do not add up, and when figures do not add up, I cannot sleep.

I have written on the sheet, in large letters, as though capitals might make the truth clearer:

THEORETICAL CALCULATION: 9 systems × 0.3W (declared standby) = 2.7W 2.7W × 24h = 64.8 Wh = 0.065 kWh

MEASURED CALCULATION (without router): 2.7W × 24h = 64.8 Wh = 0.065 kWh

The figures coincide. Perfectly. Too perfectly, perhaps. Nine LED systems consume exactly two point seven watts on standby, which means exactly zero point three watts per system, which means exactly what the manufacturers declare in their specifications. I ought to be satisfied. I ought to feel a kind of validation — I measured, I confirmed the specifications are correct, the experiment was a success. Instead I feel a diffuse frustration, a dissatisfaction without a clear object.

Because I did not run this experiment to confirm that manufacturers tell the truth. I ran it because I suspected they were lying. Because for months I had felt there was an invisible leak in this flat, a hidden consumption I was paying for without understanding it. And now, after obsessively measuring for twenty-four hours, after disrupting the whole family and losing a night's sleep, I discover that… everything is normal? That the bulbs consume exactly what they should? That there is no conspiracy, no lie, no concealment?

I cannot accept that. I refuse to accept that the whole obsession was pointless, all my suspicions mere unfounded paranoia. Because I know — through intuition, through experience, through those warm transformers I have touched — that something is not right. That somewhere in the calculation, something is missing. That the truth cannot be this simple.

I stand up. I go to the meter. It is still connected, though the experiment is over. The screen now shows 0.130 kWh / 2.8 W. It has risen by two milliwatts. In ten minutes. Why? Because transformers warm up. Because after I disconnected and reconnected the router, I probably created a small disturbance in the electrical network, a fluctuation that made the transformers draw fractionally more, temporarily. Or perhaps it is simply normal variance. Perhaps the consumption oscillates between two point seven and three watts, depending on temperature, on humidity, on the position of the moon.

But two point eight is more than two point seven. And the difference, though small, means something. It means that if I were to measure for a week, I would probably get a slightly higher average. And if I were to measure for a year, I would probably discover that consumption rises in summer — when the transformers heat up more in the ambient warmth — and falls in winter. And that the real annual average is not two point seven but, say, three point one. Which would mean:

3.1W × 24h × 365 days = 27,156 Wh = 27.2 kWh per year = RON 19

Nineteen lei instead of sixteen. A difference of three lei. Three lei that appear in no theoretical calculation, mentioned in no specification, but which exist. Somewhere. In the real friction of the physical world, in the imperfections of components, in the slow degradation of capacitors that have now been running for two, three, four years.

I think about that: degradation. My LED systems were installed at different times, over the years. The hallway ones have four years on them. The living room, three. The bathroom, three. The bedroom, two. The kitchen, two. And throughout all that time, their transformers have been permanently connected to the mains. They have sat in their sockets, in their plastic boxes, consuming — day and night, summer and winter, including the two weeks every year when we go away and the flat stands empty. The transformers do not know about holidays. They do not take breaks. They function, always, as long as they are plugged in.

And electronic components degrade. That is a fact. Electrolytic capacitors — the most vulnerable component in any transformer — lose capacity over time. Their ESR — equivalent series resistance — rises. Efficiency falls. For a new capacitor, ESR might be ten milliohms. After two years of continuous operation, perhaps fifty. After four years, perhaps a hundred. And the higher the ESR, the more energy is lost as heat. The more the transformer consumes to do the same job: keeping the circuit in standby, ready to receive the WiFi command.

I make a rough estimate. If new transformers draw 0.3W per system, and degradation increases consumption by ten per cent per year — a conservative estimate, based on data about capacitor ageing — then:

Year 1: 0.30WYear 2: 0.33WYear 3: 0.36WYear 4: 0.40W

And my systems are different ages. The hallway ones, four years old, are probably drawing around 0.40W each. The living-room ones, three years old, around 0.36W. The weighted average would be somewhere around 0.33 to 0.34W per system. Which means:

9 × 0.34W = 3.06W total

Three point zero six. More than the two point seven I measured. Why the discrepancy? Because I measured twenty-four hours. One day. A single moment in the lives of transformers that have been running for thousands of days. And in a single day, variations are small. But over a year, degradation becomes visible. Consumption rises. And nobody measures it. Nobody checks. Because who has the patience to measure for an entire year? Who keeps a meter connected for three hundred and sixty-five days?

I sit back down. I feel the frustration transforming into something more concrete: a sad understanding of the limits of empirical knowledge. I ran an experiment. I measured something. But what did I measure? A snapshot. A static photograph of a dynamic process. I measured the consumption of the lights on the eighteenth of April 2024, at an ambient temperature of approximately nineteen degrees Celsius, with the transformers in their current state of degradation.

PROVISIONAL CONCLUSION:Lights consume approximately 2.7–3.0W on standby.Annual cost: RON 17–20.Uncertainty: ±15% due to degradation and variation.

The word "provisional" bothers me. The word "approximately" bothers me. The word "uncertainty" bothers me. Because I want definitive truth, not provisional truth. I want exact figures, not approximations. I want certainty, not uncertainty.

But I cannot have them. Because the world does not operate on absolute certainties. The world operates on approximations good enough to make decisions by. And two point seven watts is a good enough approximation.

I must make peace with that. Even if I cannot.

V. THE INVESTIGATION — Looking for explanations

The anatomy of the nine systems: what they actually are

The individual measurements the next morning reveal the full picture. After methodically connecting each system alone and waiting for the reading to stabilise:

HALLWAY LEFT: 0.22WHALLWAY RIGHT: 0.23WLIVING ROOM CENTRE: 0.41WBEDROOM LAMP LEFT: 0.28WBEDROOM LAMP RIGHT: 0.28WKITCHEN LED STRIP: 0.88WBATHROOM MIRROR LEFT: 0.24WBATHROOM MIRROR RIGHT: 0.25WBATHROOM SHOWER: 0.23W

Total: 0.22 + 0.23 + 0.41 + 0.28 + 0.28 + 0.88 + 0.24 + 0.25 + 0.23 = 3.02 W.

There is the culprit. The kitchen LED strip's transformer — the big one, with its compulsively blinking blue LED — draws 0.88W on standby. Nearly a full watt. Almost four times as much as the simple hallway spots. Almost twice as much as the complex dimmer fitting in the living room.

The reason is clear once I think about it. The strip feeds fifty-four LEDs. The transformer must be capable of delivering high current when the strip is fully lit. Larger transformers have larger components — larger capacitors, larger inductors, more complex regulation circuits. All of these consume more on standby, because losses are proportional to component size. And there is that blue LED, blinking every two seconds. The LED itself draws perhaps a milliwatt. But the circuit controlling the blink — the one opening and closing a transistor dozens of times a minute — draws far more. A small aesthetic luxury, to show that the device is on standby, that it is functional, ready to receive my command. It costs approximately seven lei per year. Seven lei for a blue eye blinking inside a closed cupboard that nobody looks into except when fetching washing-up liquid.

I touch the transformer before disconnecting it. After only ten minutes on standby, it is already hot — not burningly so, but uncomfortable. Left connected for a full day, it would probably stabilise at around fifty degrees Celsius. And at elevated temperatures, components degrade exponentially faster. A capacitor that at twenty degrees would last ten years will last only two years at fifty. Another vicious circle: high consumption produces high heat; high heat accelerates degradation; degradation increases consumption; increased consumption produces more heat.

Power factor and the mathematics I hadn't been taught

There is still the question of why the sum of individual measurements — 3.02W — differs from the twenty-four-hour experiment result of 2.7W. The answer, I discover, lies in power factor.

Power factor — expressed as cos φ — is the ratio between active power (the power doing useful work) and apparent power (the power flowing through the circuit). For a perfect device, power factor is one. For cheap switching transformers, especially on standby, it can be as poor as 0.3 or 0.4.

What this means in practice: when my meter shows 2.7 watts, it is measuring active power — the power that maintains the standby circuits plus the heat losses. But more power than 2.7W is actually flowing through the household wiring. The apparent power is higher. How much higher depends on the power factor.

If power factor is 0.7, then apparent power = 2.7W / 0.7 = 3.86 VA (volt-amperes). The difference between 3.86 VA and 2.7W is reactive power — current that oscillates back and forth without transferring net energy, but which nonetheless exists, loads the wiring, and generates cable losses.

This also explains the discrepancy between individual and combined measurements. When measuring systems individually, I measure each in isolation. When they all run together, their reactive power components interact with the network in complex, non-linear ways. The combined active-power reading of 2.7W differs from the sum of individual readings of 3.02W because of these interaction effects.

I am left in a state of incomplete knowledge. I know that nine systems consume approximately 3W on standby when measured individually. I know they consume approximately 2.7W when measured together. I know the difference is probably due to power factor and non-linear interactions. But I do not know exactly how much of the 2.7W represents truly useful active power and how much is measurement artefact, distortion, effects I cannot fully understand without professional-grade instruments.

And perhaps that is the lesson. Perhaps my obsession with measuring everything, reducing uncertainty to zero, is doomed — not because I am not trying hard enough, but because reality itself is irreducibly uncertain. Every measurement is a simplification. Every figure is an abstraction. Every understanding is incomplete.

But nonetheless. Three watts. Three watts that represent approximately twenty lei per year. And of those three, almost one watt — nearly a third — comes from the kitchen LED strip. From that hot, inefficient transformer, blinking its blue eye inside a closed cupboard, consuming energy to produce information that nobody sees.

Degradation: the invisible obsolescence

The manufacturers know this. Of course they do. They design transformers with capacitors whose lifespan is calculated — not infinite, not maximal, but calculated to last a little beyond the warranty period. Because if you fitted premium capacitors rated for twenty years, the transformer would cost three lei more. And the consumer, standing in a shop comparing two apparently identical transformers — one at twenty lei, one at twenty-three — will choose the cheaper one. Because those three extra lei are visible, concrete, immediate. But the savings from long-term efficiency are invisible, abstract, spread across twenty years of use.

So the manufacturer optimises for price rather than durability. You buy cheaply now, but pay more in the long run. And after four or five years, when the transformer has become significantly less efficient, you don't replace it. Because it still works. It still lights the LED. It still responds to commands. It just consumes more. But how would you know? You are not sitting there with a meter. You are not measuring. You are not comparing with how it was when it was new.

So you pay. Year after year. The bill rises slowly, imperceptibly. You assume it is because you ran the air conditioning more this summer. Or because tariffs went up — and they probably did, but not by that much. Part of the increase is simply the slow degradation of your electrical infrastructure. Invisible wear. Oxidising capacitors. Rising ESR. Falling efficiency.

And this, I realise, is the subtlest form of planned obsolescence. Not that the product breaks completely. Not that it stops working. It simply becomes gradually less efficient, and that inefficiency translates into rising costs for you — costs you pay without knowing you are paying them, without attributing them to their correct cause.

The projection

I look at the degradation projections I have drawn up, assuming five per cent annual deterioration:

LED strip (now: 0.88W):

• Year 1 (2025): 0.92W
• Year 2 (2026): 0.97W
• Year 4 (2028): 1.07W
• Year 6 (2030): 1.18W

Living-room centre (now: 0.41W):

• Year 1: 0.43W
• Year 2: 0.45W
• Year 4: 0.50W
• Year 6: 0.55W

All nine systems total (now: 3.02W):

• Year 1: 3.17W
• Year 2: 3.33W
• Year 4: 3.67W
• Year 6: 4.05W

I look at the projection. In six years, standby consumption will have risen from three watts to four. A thirty-three per cent increase. Assuming no transformer fails completely, assuming all of them continue functioning — just with mounting inefficiency.

Four watts in 2030. Thirty-five kilowatt-hours per year. Twenty-five lei per year. An increase of nine lei from now, simply from degradation — without adding a single new bulb, without changing anything in my behaviour.

Nine lei. More than a new transformer for the LED strip costs. So it would be economically efficient to replace it now, preventively, before it degrades further. But I will not do that. Because nine lei a year is still not enough to motivate me. Because replacement means ordering the transformer — finding the exact model, searching online, checking specifications — then waiting for delivery, then dismantling the strip, disconnecting the old transformer, connecting the new one, verifying it works. Hours of work. For nine lei a year.

It is not worth it. In purely economic terms, it is not worth it. My time is worth more than the potential savings. So I will let the transformer degrade. I will pay the extra nine lei. And I will contribute to the culture of the passive consumer who accepts inefficiency as inevitability, who pays for degradation without protesting, without acting, without asking for anything better.

And the manufacturers know this. They calculate this balance precisely: degradation must be slow enough not to be perceived as an immediate problem, but fast enough to generate additional revenue across the product's lifetime. It is a fine equilibrium, optimised through millions of transactions, through consumer behaviour statistics, through sophisticated economic models that predict exactly how much a product can degrade before the consumer revolts and demands a replacement.

For transformers, this equilibrium is perfect. The degradation is slow enough — five per cent per year — that nobody notices. It is buried in the noise of bills that vary anyway because of seasonality, because tariffs change, because behaviour fluctuates. Nobody looks at the January 2024 bill and compares it with January 2020 and says: "I'm consuming ten per cent more though my behaviour is the same — my transformers must have degraded." Nobody does that. Because it is absurdly complicated. Because it requires a perfect memory of past behaviour. Because it presupposes all other variables have remained constant, which they never do.

So the degradation remains invisible. It sits in that grey zone between the observable and the unobservable, between the significant and the negligible. And you pay. Year after year. Without knowing. Without being able to prove it. Without anyone to blame, because nobody promised your transformer would remain efficient for ever. They only promised it would work for two years — the warranty period. And it does work. It just works progressively worse.

CLOSING — The darkness that teaches, the darkness that traps

I get up from the table. It is one o'clock. I must get back to the hospital website. I must finish the appointments module by Friday. I must do productive things, things that generate income, things that justify my professional existence.

But before that, I do one last thing: I write on the sheet, below the projection, a note to my future self:

JANUARY 2025 — RE-EVALUATE:Measure the LED strip again. If >1W, replace the transformer.Check whether you have done this. You probably haven't.Accept that you are weak. Accept that you pay for weakness.

I look at the note. It is brutally honest. It is a prediction of my own failure. But at least it is honest. And honesty, even in the anticipation of failure, is better than the illusion that I will do things correctly.

I leave the sheet on the table. I go to my desk. I open the code editor. I look at the lines of PHP that need writing, the validations that need implementing, the tests that need running. I try to concentrate. I try to leave behind the obsession with transformers, with degradation, with the three watts consuming in silence while I try to work.

But in my mind, the image remains: nine black boxes dispersed through the flat, each slightly warm, each degrading slowly, each consuming fractionally more than yesterday, each contributing to a slow, invisible, inevitable rise in my electricity bill.

And the fridge hums in the kitchen. And the house breathes electrically. And I write code, knowing that as I write, the capacitors are oxidising, the ESR is rising, the efficiency is falling.

And life continues. Because that is what life does: it continues, regardless of our obsessions, regardless of our measurements, regardless of our knowledge of inevitable degradation.

It continues. And we pay. For electricity. For degradation. For the illusion that we have control.

And perhaps that is all we can do: pay, know that we are paying, and carry on. Because the alternative — complete disconnection, the abandonment of everything convenient, the return to simplicity — is too radical, too alienating, too impossible for someone who has grown accustomed to instant light, to warmth at the press of a button, to the perpetual comfort that costs only three watts and about which one need never think.

Until you think. And then you cannot stop thinking. And the obsession begins. And you measure. And you learn. And the knowledge does not liberate you — it buries you deeper in your own inability to change anything.

And that is the saddest thing the entire experiment taught me: that there are things it is better not to know. Because once you know them, you cannot forget them. And they will follow you. In every month when the bill arrives. In every moment when you press a button and the light comes on instantly. In every night when you lie down knowing that in the dark, your transformers are consuming, degrading, costing you.

And you can do nothing. Except accept. And pay. And continue.